Article (PDF Available)

Altering misperception of sleep in insomnia: Behavioral experiment versus verbal feedback

Department of Psychology, University of California, Berkeley, Berkeley, California, United States
Journal of Consulting and Clinical Psychology (Impact Factor: 4.85). 09/2006; 74(4):767-76. DOI: 10.1037/0022-006X.74.4.767
Source: PubMed
ABSTRACT
Forty-eight individuals with insomnia were asked to wear an actigraph and keep a sleep diary for 2 nights. On the following day, half were shown the discrepancy between the data recorded on the actigraph and their sleep diary via a behavioral experiment, whereas the other half were told of the discrepancy verbally. Participants were then asked to monitor their sleep for 2 further nights to index the effect of these interventions. Although both reduced sleep misperception, the behavioral experiment (effect size: 0.79 to 1.25) led to greater reduction in self-reported sleep impairment, insomnia symptoms, and sleep-related anxiety and distress compared with verbal feedback (effect size: -0.06 to 0.31). Further, the patients regarded the behavioral experiment as a more beneficial and acceptable intervention strategy than verbal feedback.

Full-text (PDF)

Available from: Nicole K Y Tang
Altering Misperception of Sleep in Insomnia:
Behavioral Experiment Versus Verbal Feedback
Nicole K. Y. Tang
King’s College London
Allison G. Harvey
University of California, Berkeley
Forty-eight individuals with insomnia were asked to wear an actigraph and keep a sleep diary for 2 nights.
On the following day, half were shown the discrepancy between the data recorded on the actigraph and
their sleep diary via a behavioral experiment, whereas the other half were told of the discrepancy
verbally. Participants were then asked to monitor their sleep for 2 further nights to index the effect of
these interventions. Although both reduced sleep misperception, the behavioral experiment (effect size:
0.79 to 1.25) led to greater reduction in self-reported sleep impairment, insomnia symptoms, and
sleep-related anxiety and distress compared with verbal feedback (effect size: 0.06 to 0.31). Further, the
patients regarded the behavioral experiment as a more beneficial and acceptable intervention strategy
than verbal feedback.
Keywords: behavioral experiment, insomnia, misperception, anxiety, treatment acceptability
Insomnia is among the most prevalent psychological health
problems, with at least 1 in 10 people suffering from insomnia
chronically (Simon & Von Korff, 1997). It has serious conse-
quences, including impaired cognitive function, functional impair-
ment, work absenteeism, increased use of medical services (Roth
& Ancoli-Israel, 1999), and doubling the risk of accident (Ohayon,
Caulet, Priest, & Guilleminault, 1997). Moreover, longitudinal
studies indicate that insomnia significantly heightens the risk of
developing first-onset depression, an anxiety disorder, and sub-
stance dependency (e.g., Ford & Kamerow, 1989). Not surpris-
ingly, then, the costs to society are enormous. In the United States,
the direct and indirect costs associated with insomnia were con-
servatively estimated to be between $92.5 and $107.5 billion per
year (Stoller, 1994).
There has been significant progress in the development of
cognitive– behavioral treatments for insomnia (Morin et al., 1999).
However, the field is not yet at a point where patients can be
offered a maximally effective treatment. This is indicated by the
significant proportion (19%–26%) of patients who do not improve
following cognitive– behavioral treatments and by the average
overall improvement among those who do respond being approx-
imately 50%–60% (Morin, Culbert, & Schwartz, 1994; Murtagh &
Greenwood, 1995). Although this amount of change is statistically
significant, it is not enough to ensure that the average patient
becomes a “good sleeper” (Harvey & Tang, 2003). One important
feature of contemporary cognitive– behavioral treatments that has
not as yet been incorporated into psychological treatments for
insomnia is the use of behavioral experiments.
Behavioral Experiments
A behavioral experiment is a planned experiential activity that is
undertaken by patients in or between therapy sessions. A behav-
ioral experiment is individually tailored and developed collabora-
tively with the patient. The ultimate aim of a behavioral experi-
ment is to give the patient the opportunity to experimentally test
out their beliefs and behaviors, and by so doing, it provides the
patient with clear and memorable demonstrations that their beliefs
or behaviors are unhelpful or maladaptive (e.g., Clark, 1999;
Salkovskis, 1991). Although behavioral experiments were first
developed as an alternative to the conventional verbal techniques
used to challenge cognitive distortions in depression and anxiety
disorders (e.g., A. T. Beck, Rush, Shaw, & Emery, 1979), they are
increasingly included in cognitive– behavioral treatments for a
range of other psychological problems (Bennett-Levy et al., 2004).
Compared with conventional verbal techniques, a behavioral ex-
periment is thought to be a more powerful treatment strategy for
promoting change (e.g., J. S. A. Beck, 1995; Greenberger &
Padesky, 1995). However, this assumption has been tested only to
a limited extent (e.g., Harvey, Clark, Ehlers, & Rapee, 2000;
Salkovskis, Clark, Hackmann, Wells, & Gelder, 1999). In partic-
ular, the evidence base for the application of behavioral experi-
ments to psychological problems outside the anxiety disorders
remains weak.
Altering Misperception of Sleep in Insomnia
In the context of insomnia, a substantial proportion of patients
exhibit a tendency to misperceive the quantity of their sleep (e.g.,
Carskadon et al., 1976; Perlis, Giles, Mendelson, Bootzin, &
Wyatt, 1997). Specifically, they tend to overestimate the time they
take to get to sleep (sleep onset latency) and underestimate the
total amount of time they have slept (total sleep time), relative to
Nicole K. Y. Tang, Department of Psychology, Institute of Psychiatry,
King’s College London, London, England; Allison G. Harvey, Psychology
Department, University of California, Berkeley.
This research was supported by Grant 065913 from the Wellcome Trust
and Grant R00023853 from the Economic and Social Research Council.
Correspondence concerning this article should be addressed to Nicole
K. Y. Tang, Department of Psychology, Institute of Psychiatry, King’s
College London, De Crespigny Park, London SE5 8AF, England. E-mail:
n.tang@iop.kcl.ac.uk
Journal of Consulting and Clinical Psychology Copyright 2006 by the American Psychological Association
2006, Vol. 74, No. 4, 767–776 0022-006X/06/$12.00 DOI: 10.1037/0022-006X.74.4.767
767
Page 1
normal sleepers (e.g., Frankel, Buchbinder, Coursey, & Snyder,
1976; Perlis, Smith, Andrews, Orff, & Giles, 2001). A number of
theoretical models have been proposed in an attempt to account for
these findings (e.g., Borkovec, 1982; Harvey, 2002; Lundh &
Broman, 2000; Perlis et al., 1997). There is experimental evidence
supporting a role for physiological arousal (Bonnet & Arand,
1994, 1996), cognitive arousal (Tang & Harvey, 2004b), and brief
arousals of 3 to 30 s (Smith & Trinder, 2000). Moreover, there is
correlational evidence supporting a role for cognitive arousal (Van
Egeren, Haynes, Franzen, & Hamilton, 1983) and high-frequency
electroencephalogram activity in the beta (14 –35 Hz) range (Perlis
et al., 2001).
Functionally, misperception of sleep has been proposed to be
one of several cognitive processes that contribute to the mainte-
nance of insomnia (Harvey, 2002). This viewpoint suggests that
misperception of sleep has at least two adverse consequences:
First, it leads the person to believe that he or she is not getting
sufficient sleep, which in turn fuels worry and anxiety about sleep.
Second, given that excessive worry and escalating anxiety are
conditions that are antithetical to sleep, misperception of sleep may
end up contributing directly to the sleep disturbance (Harvey,
2002). Consistent with the first of these proposals, Neitzert-Semler
and Harvey (2005) demonstrated that experimentally manipulating
the perception of the quality of sleep immediately on waking
caused participants given feedback that their sleep was of poor
quality to worry more and feel more anxious about their sleep
during the subsequent day. Accordingly, it may be fruitful to target
misperception of sleep while treating individuals with insomnia.
Two previous studies have demonstrated that it is possible to
correct misperception of sleep among individuals with insomnia.
In the first, Downey and Bonnet (1992) woke patients 27 times per
night and provided them with polysomnography (PSG) feedback
about whether they were asleep or awake. This intervention was
effective. To capitalize on this result, while minimizing the intru-
siveness to the patient and maximizing utility to clinicians (most of
whom do not have access to PSG), we devised a novel behavioral
experiment (Tang & Harvey, 2004a). This involved vividly dem-
onstrating the discrepancy between subjectively estimated sleep
(drawn from a daily sleep diary) and objectively estimated sleep
(drawn from a wrist-worn actigraph). In a pilot study of this
method, we found that, compared with individuals who were not
shown the discrepancy, those shown the discrepancy estimated
their sleep onset latency to be shorter and reported less anxiety and
preoccupation about their sleep following the behavioral experi-
ment. However, there were several key limitations of this study
that restrict the conclusions that can be drawn. First, although the
study demonstrated that individuals who engaged in the behavioral
experiment experienced greater therapeutic benefits compared
with those who did not, the relative effectiveness of altering
perception of sleep using a behavioral experiment versus verbal
techniques (i.e., education about likely misperception) is yet to be
determined. Second, apart from sleep estimates, the study had only
one outcome measure (sleep-related anxiety and preoccupation). It
is of clinical interest to investigate the effect of the behavioral
experiment on a range of outcomes. Third, a student sample who
met diagnostic criteria for insomnia was used. Hence, replication
with a more representative sample outside the university popula-
tion is required.
The Present Study
The aim of the present study was to compare the therapeutic
effect of showing the discrepancy between subjective and objec-
tive sleep estimates (a behavioral experiment) with the effect of
simply being told about the discrepancy (verbal explanation). The
experimental protocol of Tang and Harvey (2004a) was altered in
five ways: (a) The participants were individuals with insomnia
recruited from the general public so as to increase the generaliz-
ability of the findings; (b) two standardized questionnaires and two
rating scales were added to index the impact of the behavioral
experiment on a broader range of outcomes; (c) the length of the
baseline period and the experimental period was shortened from 3
to 2 nights to enable the experiment to be completed on weekdays
and within 1 week, minimizing the variability introduced by par-
ticipants adopting a different sleep pattern over the weekend; (d)
four postexperiment rating scales were administered to evaluate
the acceptability of the intervention; and (e) a control group was
included to compare the effects of the novel behavioral experiment
with those of conventional verbal techniques. On the basis of
previous theoretical and empirical work (e.g., Bennett-Levy et al.,
2004; Harvey et al., 2000; Salkovskis et al., 1999) it was predicted
that, relative to participants who were simply told of the discrep-
ancy, participants who were shown the discrepancy via the behav-
ioral experiment would (a) think and feel more positively about
their sleep and (b) estimate their sleep more accurately.
Method
Participants
Participants were recruited from the community through posters placed
around the city asking individuals aged 18 60 who had “trouble with
sleeping” to contact the experimenter. Participants admitted to the study
had to have (a) met strict criteria for primary insomnia as provided in the
Diagnostic and Statistical Manual of Mental Disorders (4th ed., text rev.
[DSM–IV–TR]; American Psychiatric Association, 2000), (b) experienced
sleep disturbance at least 3 nights per week over the past month (World
Health Organization, 1992), and (c) scored above 5 on the Pittsburgh Sleep
Quality Index (Buysse, Reynolds, Monk, Berman, & Kupfer, 1989). Of the
101 people who responded to the recruitment drive, 42 were not included
owing to lack of interest or time. A further 11 were excluded following the
diagnostic screening for not meeting the full criteria for insomnia (n 6),
because the sleep disturbance was secondary to another psychological
disorder (n 1) or another sleep disorder (n 1), for the current use of
psychotropic prescriptions (n 1), or for the regular use of recreational
drugs (n 2). Thus, the final sample comprised 48 individuals with
primary insomnia.
Materials
Interview for the Diagnosis of Insomnia (IDI). We administered the
IDI (Harvey et al., n.d.) to identify the presence of insomnia in the past
month. This semistructured interview contains five sections that carefully
assess each of the DSM–IV–TR criteria for primary insomnia. It has good
internal consistency (␣⫽.87), sensitivity (92%), specificity (89%), and
test–retest reliability (r .90), as well as high diagnostic agreement for the
presence (90%) and absence (92%) of insomnia.
Pittsburgh Sleep Quality Index (PSQI; Buysse et al., 1989). As a
diagnostic check we asked the participants to complete the PSQI with
reference to sleep obtained in the past month. The PSQI is a 19-item
self-administered questionnaire, generating a global score ranging from 0
768
TANG AND HARVEY
Page 2
to 21 (with higher scores indicative of worse sleep quality). The PSQI has
good internal consistency (␣⫽.83) and test–retest reliability (r .85)
(Carpenter & Andrykowski, 1998). The cutoff of 5 identifies a clinically
significant sleep disturbance with 89.6% sensitivity and 86.5% specificity
(Buysse et al., 1989).
Sleep diary. To index subjective sleep immediately on waking, the
participants were asked to record their bedtime, sleep onset latency (SOL;
time from lights-out to sleep onset), wake time after sleep onset (total
amount of time spent awake from sleep onset to final awakening), total
sleep time (TST; total amount of sleep obtained), and the time waking up
in the morning.
Actigraph. A Mini-Motionlogger Actigraph Basic (Ambulatory Mon-
itoring, Ardsley, NY) was used to provide an objective estimate of sleep.
Cased within the actigraph is a miniaturized piezoelectric sensor pro-
grammed to detect and store physical motion data in zero-crossing mode at
60-s intervals. Stored data were downloaded for analysis using compatible
software (Action W), and conversion of data into sleep parameters was
completed using the Cole–Kripke algorithm (Cole, Kripke, Gruen, Mul-
laney, & Gillin, 1992), with Webster’s rescoring rules (Webster, Kripke,
Messin, Mullaney, & Wyborney, 1982). To facilitate more accurate scoring
of sleep, the participants were instructed to depress the event marker on the
Mini-Motionlogger once just prior to going to sleep (i.e., lights-off) and
once immediately upon waking the following morning.
Actigraphy is a useful and nonintrusive instrument to measure the
sleep–wake schedule (DeSouza et al., 2003). Relative to PSG sleep esti-
mates, the gold standard, the epoch-by-epoch agreement rate for sleep and
wakefulness detection in adult good sleepers ranges from 74% to 98%
(Ancoli-Israel, Clopton, Klauber, Fell, & Mason, 1997; Reid & Dawson,
1999; Sadeh, Sharkey, & Carskadon, 1994). The correlation between
actigraphic and PSG estimates ranges from .77 to .98 for SOL and from .82
to .90 for TST in adult good sleepers (Cole et al., 1992; Sadeh et al., 1994).
Actigraphy is highly sensitive (sensitivity: 87%–99%) at detecting sleep
epochs identified by PSG; it may, however, be less reliable at detecting
wake and sleep onset (specificity: 28%–90%; Kuishida et al., 2001; Reid &
Dawson, 1999; Sadeh, Hauri, Kripke, & Lavie, 1995), especially in indi-
viduals with insomnia or depression, who may be able to lie immobile for
long periods. Although we recognized this limitation of actigraphy, we
chose to utilize this technology for the behavioral experiment because (a)
it permitted the sleep of the patients to be estimated in a natural sleeping
environment in an unobtrusive way, (b) its capacity to produce instanta-
neous analysis of sleep data enabled the experimenter to provide the
patients timely feedback within the session, and (c) its impressive graphical
and statistical display of sleep data allowed the patients, regardless of their
educational background, to receive unambiguous feedback regarding their
sleep.
Baseline measures. Three validated questionnaires were used to index
baseline anxiety, depression, and worry. They are described as follows:
The Beck Anxiety Inventory (BAI; A. T. Beck, Epstein, Brown, & Steer,
1988) is a 21-item scale that measures common symptoms of anxiety over
the past week. The items are summed to obtain a total score that ranges
from 0 to 63 (with higher scores indicative of greater anxiety). The BAI has
good internal consistency (␣⫽.92) and test–retest reliability (r .75).
The Beck Depression Inventory (BDI; A. T. Beck, Ward, Mendelson,
Mock, & Erbaugh, 1961) is a 21-item scale that measures common symp-
toms of depression in the past month. The items are summed to obtain a
total score that ranges from 0 to 63 (with higher scores indicative of more
severe depression). The BDI has well-established internal consistency
(split-half Pearson r .93) and validity, as indicated by the strong
correlation between the BDI score and clinicians’ ratings of depth of
depression (mean r .66).
The Penn State Worry Questionnaire (PSWQ; Meyer, Miller, Metzger,
& Borkovec, 1990) is a 16-item scale that measures the trait of worry.
The items are summed to obtain a total score that ranges from 16 to 80
(with higher scores indicative of greater tendency to worry). The PSWQ
has good internal consistency (␣⫽.91–.95) and high test–retest reli-
ability (r .74).
Postfeedback measures. Three self-report questionnaires and two rat-
ing scales were used to measure the participants’ change in response
associated with the feedback. They are described as follows:
The Anxiety and Preoccupation About Sleep Questionnaire (APSQ;
Tang & Harvey, 2004a) is a 10-item measure that assesses sleep-related
anxiety and preoccupation associated with insomnia. The participants were
asked to rate, on a 10-point scale, how true each of the statements was for
them over the past 2 days (1 not true,10 very true). The items were
summed to obtain a total score that ranged from 10 to 100 (with higher
scores indicative of higher level of anxiety and preoccupation about sleep).
The APSQ has high internal consistency (␣⫽.92). The high correlations
between the APSQ and the PSQI (r .44, p .001) and BAI (r .37, p
.001; A. T. Beck et al., 1988) indicate that higher scorers on the APSQ are
also higher scorers on the PSQI and BAI.
The Insomnia Symptom Questionnaire (ISQ; Spielman, Saskin, &
Thorpy, 1987) is a 13-item measure that assesses the occurrence of insom-
nia symptoms. The original categorical response scale to each item (1
rarely,2 sometimes,3 frequently) was replaced by a 10-cm visual
analogue scale, with the left end of the continuum labeled never (score
0) and the right end always (score 10). The participants were asked to
respond to each item with reference to the last 2 days. The total score of the
ISQ ranges from 0 to 130 (with lower scores indicative of better sleep or
functioning). The ISQ has adequate internal consistency (␣⫽.73) and is
a useful tool to index pre–post treatment reduction in insomnia symptoms.
The Insomnia Severity Index (ISI; Bastien, Vallieres, & Morin, 2001) is
a seven-item measure of the overall degree of improvement in sleep. The
participants were asked to complete the ISI with reference to the last 2
days. The total score of the ISI ranges from 0 to 28 (with higher scores
indicative of more severe insomnia). The ISI has adequate internal and
concurrent validity (␣⫽.76 –.78; correlation with sleep diary variables
.32–.91; correlation with PSG variables .07–.45) and has been shown to
be sensitive to improvement in sleep over the course of treatment for
insomnia.
The sleep-perception rating was a single-item scale used to assess
change in the way the participants thought about their sleep over the last 2
days: “Overall, how well do you think you sleep?” (response scale: 0 not
very well,10 very well).
The sleep-distress rating was a single-item scale used to assess change
in the way the participants felt about their sleep over the last 2 days: “How
distressed are you by your sleep problem?” (response scale: 0 not at all,
10 very much).
Postexperiment ratings. At the end of the experiment, we asked the
participants to give ratings, on four 11-point scales, of the acceptability of
the feedback procedures: (a) “How enjoyable was this study?” (response
scale: 0 not at all enjoyable,10 very enjoyable); (b) “To what extent,
if at all, did you benefit from participating in this study?” (response scale:
0 I benefited nothing at all,10 I benefited a lot); (c) “How acceptable
did you consider this procedure for your insomnia?” (response scale: 0
not at all acceptable,10 very acceptable); and (d) “How acceptable
would you consider this procedure for other people with insomnia?”
(response scale: 0 not at all acceptable,10 very acceptable).
Design and Procedure
The protocol of this study was approved by the Research Ethics Com-
mittee, University of Oxford, England. Participants with insomnia were
randomly assigned to either the behavioral experiment group (n 24) or
the verbal feedback group (n 24). Each participant attended three
sessions.
Session 1. After obtaining informed consent, we administered the IDI.
The participants then completed the PSQI, BAI, BDI, and PSWQ. To index
baseline misperception, the participants were asked to wear an actigraph
769
BEHAVIORAL EXPERIMENT VERSUS VERBAL FEEDBACK
Page 3
and keep a sleep diary for 2 consecutive days (Days 1–2). To counteract
expectation effects, the participants were told that the purpose of the study
was “to observe the sleep–wake schedule over a 4-day period” and that
they would need to return to the lab after completing the first 2 nights
(Days 1–2) of sleep recording “to download the data and to renew the
battery of the actigraph” such that they could complete 2 further nights of
sleep recording (Days 3–4). A second session was thus scheduled on Day
3. Before leaving the session, the participants were reminded to maintain
their normal sleep–wake patterns and to refrain from using sleep medica-
tions and excessive alcohol throughout the experiment.
Session 2. When the participants returned on Day 3 (the feedback
session), the actigraph and the sleep diaries for Days 1–2 were collected.
The participants then completed the APSQ, ISQ, and ISI (with reference to
their sleep patterns over the last 2 days) to provide an index of prefeedback
sleep-related anxiety and preoccupation, occurrence of insomnia symp-
toms, and overall sleep impairment. Before receiving the feedback, the
participants were also asked to give a sleep-perception rating and a sleep-
distress rating. The participants in the behavioral experiment group then
participated in the behavioral experiment, which involved (a) the partici-
pants being shown and taught to read their actigraph recordings from the
last 2 days and (b) the participants completing a data summary sheet on
which the sleep diary and the actigraph recordings were juxtaposed. The
participants then calculated the discrepancy between the sleep diary data
and the actigraphic sleep estimates. The aim of this procedure was to
provide a clear and memorable demonstration of the discrepancy between
their self-reported sleep and objectively estimated sleep. The verbal feed-
back group were given the same information verbally rather than behav-
iorally. The procedure for delivering the verbal feedback was that the
experimenter downloaded the sleep data from the actigraph to a computer
while the participant waited in the interview room for about 25 min
(matching the length of the behavioral experiment). During this time, the
experimenter performed a brief analysis of the participant’s sleep data in the
adjacent office. When the brief analysis was finished, the experimenter ver-
bally explained to the participants whether the data indicated that they had
overestimated the time they had taken to get to sleep and whether they had
underestimated the amount of sleep they had obtained. After receiving feed-
back, the participants were instructed to wear the actigraph for the next 2 nights
and to keep a sleep diary for the next 2 mornings (Days 3– 4). These additional
2 nights were added to index the impact of the feedback session.
Session 3. During the final session, the actigraph and the sleep diaries
for Days 3 to 4 were collected. All participants were asked to complete the
APSQ, ISQ, and ISI (with reference to their sleep patterns over the last 2
days) to provide an index of their postfeedback sleep-related anxiety and
preoccupation, occurrence of insomnia symptoms, and overall sleep im-
pairment. In addition, the participants gave a second sleep-perception
rating and a second sleep-distress rating to index how well they thought
they had slept and how distressed they felt over the 2 days following the
feedback session. The participants were then debriefed. They were told that
altering the perception of sleep might help some people with insomnia to
sleep better and that this experiment actually sought to compare the
effectiveness of two procedures for altering the perception of sleep. After
the debriefing, the participants were asked to give postexperiment ratings
regarding the acceptability of the procedures in which they had participated
over the past 4 days. Finally, the participants were paid an honorarium.
Statistical Analysis for Major Outcome Variables
Unless otherwise specified, the analyses for the major outcome variables
involved a series of repeated measures analyses of variance with group
(behavioral experiment group vs. verbal feedback group) as the between-
subjects variable and session (prefeedback vs. postfeedback) as the within-
subject variable. To explore significant interactions, t tests were conducted
with a Bonferroni adjustment to control for multiple comparisons ( p
.0125). Instead of reporting all sleep variables, we focused on SOL and
TST because (a) only these two sleep parameters have consistently been
found to be related to sleep misperception in insomnia (e.g., Carskadon et
al., 1976; Frankel et al., 1976) and (b) actigraphy is less reliable in
detecting brief awakenings and hence may be less accurate in estimating
wake time after sleep onset (e.g., Sadeh et al., 1995).
Results
Participant Characteristics
Table 1 presents the participant characteristics. There were no
significant differences between the behavioral experiment group
and the verbal feedback group on sex composition (analyzed with
Table 1
Participant Characteristics
Variable
Behavioral experiment
group (n 24)
Verbal feedback
group (n 24)
2
(1, N 48)
t(46)
Sex
Female 13 14 0.09
Male 11 10
Age (years) 29.2 (11.8) 33.8 (15.9) 1.15
Typical sleep onset latency (min) 59.0 (33.7) 48.3 (31.8) 1.12
Typical total sleep time (hr) 6.1 (1.7) 6.1 (1.1) 0.14
Typical sleep quality
a
2.9 (0.6) 3.0 (0.6) 0.47
Typical sleep efficiency (%) 70.3 (18.8) 74.4 (11.1) 0.92
PSQI 11.0 (2.8) 10.7 (3.2) 0.29
BAI 11.2 (6.6) 9.3 (4.9) 1.12
BDI 11.0 (5.0) 10.8 (6.5) 0.17
PSWQ 53.3 (11.1) 46.6 (12.6) 1.97
Insomnia duration (years) 5.3 (5.6) 4.6 (4.3) 0.50
Note. Except for sex, where frequency is reported, mean values are presented with standard deviations in
parentheses. Sleep onset latency time from initial lights out to sleep onset. Total sleep time total amount
of time sleeping computed by subtracting total wake time from total time in bed. Sleep efficiency (total sleep
time/total time in bed) 100%. PSQI Pittsburgh Sleep Quality Index; BAI Beck Anxiety Inventory; BDI
Beck Depression Inventory; PSWQ Penn State Worry Questionnaire.
a
Response scale: 1 very good, 2 fairly good, 3 fairly bad, 4 very bad.
770
TANG AND HARVEY
Page 4
chi-square), age, sleep pattern over the past month (as indicated by
SOL, TST, sleep efficiency, and sleep quality extracted from the
PSQI), scores on the PSQI, BAI, BDI, and PSWQ, and insomnia
duration (analyzed with independent-samples t tests; range
0.5–20 years). The mean SOL, TST, and sleep efficiency values of
this sample over the past month were 53.6 min, 6.1 hr, and 72.4%,
respectively (as compared with the 31 min, 6.5 hr, and
85% cutoffs for insomnia; Lichstein, Durrence, Taylor, Bush, &
Riedel, 2003; Morin, 1993). Before the feedback, the participants
as a group overestimated their SOL by 29 min (range ⫽⫺24.5 to
138.5 min) and underestimated their TST by 52 min (range
206.5 to 70.5 min). These figures are comparable to those
documented by other research using PSG (Perlis et al., 1997) and
indicate that on average this sample of participants did exhibit
misperception of sleep. Consistent with the view that sleep mis-
perception occurs along a continuum (Edinger & Fins, 1995), it is
noted that the extent to which participants overestimated their SOL
and underestimated their TST varied. Some participants estimated
their sleep more or less accurately (5 min: n 2; behavioral
experiment group, n 1; verbal feedback group, n 1), and some
misperceived SOL and TST in the opposite direction (n 4;
behavioral experiment group, n 2; verbal feedback group, n
2). As the behavioral experiment was designed to specifically
target SOL overestimation and TST underestimation, these 6 in-
dividuals were excluded from the analyses below.
Degree of Overlap Among Major Outcome Measures
Correlational analyses were performed to inspect the degree of
overlap among the APSQ, ISQ, ISI, sleep-perception rating, and
sleep-distress rating using the baseline scores obtained in Session
1. Although the intercorrelations among these self-report outcome
measures were moderate to strong (see Table 2), results of reli-
ability analyses (Cronbach’s ␣⫽.16; Tukey’s test of nonadditiv-
ity, p .001) suggested that the measures should be examined
separately.
Effect of the Feedback on Questionnaire Measures and
Ratings
The mean values for the questionnaire measures and ratings
before and after the feedback are presented in Table 3. An effect
size was also calculated for each variable in this section using the
formula d (M
1
–M
2
)/ pooled. The mean value after the
feedback session was subtracted from that before the feedback
session (except for the sleep-perception rating, for which the
subtraction was reversed), such that a positive effect size denotes
an improvement in these questionnaires and ratings and a negative
effect size denotes a worsening of the symptoms measured.
For the APSQ, no significant main effect for group was ob-
served. There was a significant main effect for session, F(1, 40)
9.0, p .01, such that the average APSQ total score was lower
after the feedback session compared with before. There was also a
significant Group Session interaction, F(1, 40) 12.4, p .01.
Follow-up tests indicated that the behavioral experiment group
showed a significant reduction in their sleep-related anxiety and
preoccupation following the feedback session ( p .0125),
whereas the verbal feedback group did not ( p .69). The effect
size was 0.79 for the behavioral experiment group and 0.06 for
the verbal feedback group.
For the ISQ, no significant main effect for group was observed.
There was a significant main effect for session, F(1, 40) 18.3,
p .001, such that the average ISQ score was lower after the
feedback session compared with before. There was also a signif-
icant Group Session interaction, F(1, 40) 4.9, p .05.
Follow-up tests indicated that the behavioral experiment group
demonstrated a significant reduction in their insomnia symptoms
( p .0125). Although the verbal feedback group showed a trend
of reduction in insomnia symptoms, the pre–post feedback differ-
ence did not reach statistical significance after a Bonferroni ad-
justment ( p .03). The effect size was 1.00 for the behavioral
experiment group and 0.31 for the verbal feedback group.
For the ISI, no significant main effect for group was observed.
There was a significant main effect for session, F(1, 40) 18.9,
p .001, such that the average ISI score was lower after the
feedback session compared with before. There was also a signif-
icant Group Session interaction, F(1, 40) 4.6, p .05.
Follow-up tests indicated that the behavioral experiment group
showed a significant reduction in their sleep impairment following
the feedback session ( p .0125). Although the verbal feedback
group showed a trend of reduction in sleep impairment, the pre–
post feedback difference did not reach statistical significance after
a Bonferroni adjustment ( p .02). The effect size was 0.87 for the
behavioral experiment group and 0.27 for the verbal feedback
group.
For the sleep-perception rating, there were significant main
effects for group, F(1, 40) 5.1, p .05, and session, F(1, 40)
34.3, p .001, such that the average rating of the behavioral
experiment group was higher compared with the verbal feedback
group and that the average rating was higher after the feedback
session compared with before. Further, there was a Group
Table 2
Intercorrelations Between the Major Self-Report Outcome Measures
Variable 12345
1. APSQ
2. ISQ .62***
3. ISI .64*** .68***
4. Sleep-perception rating .37* .46** .63**
5. Sleep-distress rating .76*** .56*** .73*** .54***
Note. N 42. APSQ Anxiety and Preoccupation About Sleep Questionnaire; ISQ Insomnia Symptom
Questionnaire; ISI Insomnia Severity Index.
* p .05. ** p .01. *** p .001.
771
BEHAVIORAL EXPERIMENT VERSUS VERBAL FEEDBACK
Page 5
Session interaction, F(1, 40) 14.0, p .01. Follow-up tests
indicated that the behavioral experiment group showed a signifi-
cant increase in their sleep-perception rating following the feed-
back session ( p .0125), whereas the verbal feedback group did
not ( p .06). Moreover, after the feedback session, the sleep-
perception rating of the behavioral experiment group was signifi-
cantly higher compared with the verbal feedback group ( p
.0125). That is, whereas both groups thought that they had slept
better following the feedback session, the behavioral experiment
group had a more positive perception of their sleep relative to the
verbal feedback group. The effect size was 1.25 for the behavioral
experiment group and 0.31 for the verbal feedback group.
For the sleep-distress rating, no significant main effect for group
was observed. There was a significant main effect for session, F(1,
40) 21.2, p .001, such that the overall rating was lower after
the feedback session compared with before. There was also a
significant Group Session interaction, F(1, 40) 12.5, p .01.
Follow-up tests indicated that the behavioral experiment group
demonstrated a significant reduction in their distress about their
sleep following the session ( p .0125), whereas the verbal
Table 3
Mean Values, Standard Deviations, and Confidence Intervals for Questionnaire Measures, Ratings, Sleep Measures, and
Postexperiment Ratings
Measure
Behavioral experiment group (n 21) Verbal feedback group (n 21)
M (SD) n CI M (SD) n CI
Questionnaire measures
APSQ
Prefeedback 53.3 (18.3) 21 45.3–61.2 40.9 (17.8) 21 32.9–48.8
Postfeedback 39.1 (17.4) 21 31.0–47.3 42.0 (19.4) 21 33.9–50.1
ISQ
Prefeedback 63.3 (19.9) 21 54.3–69.3 59.8 (17.2) 21 51.6–68.0
Postfeedback 43.7 (18.9) 21 35.2–51.9 53.6 (22.0) 21 44.6–62.7
ISI
Prefeedback 14.0 (4.2) 21 11.9–16.1 13.9 (5.2) 21 11.8–16.0
Postfeedback 9.6 (5.6) 21 7.2–12.1 12.4 (5.6) 21 10.0–14.9
Ratings
Sleep-perception rating
a
Prefeedback 3.9 (1.6) 21 3.2–4.6 3.7 (1.6) 21 3.0–4.4
Postfeedback 6.1 (2.0) 21 5.3–6.9 4.2 (1.6) 21 3.4–5.0
Sleep-distress rating
b
Prefeedback 6.1 (2.3) 21 5.1–7.1 5.3 (2.3) 21 4.3–6.3
Postfeedback 4.0 (2.0) 21 3.0–4.9 5.0 (2.3) 21 4.1–6.0
Sleep measures
Subjective SOL
Prefeedback 49.2 (33.6) 20 34.5–64.0 49.2 (31.5) 20 34.4–63.9
Postfeedback 30.2 (27.0) 20 17.7–42.6 33.4 (28.0) 20 20.9–45.8
Objective SOL
Prefeedback 12.9 (7.4) 18 8.7–17.1 13.3 (9.9) 20 9.3–17.3
Postfeedback 17.1 (16.0) 18 9.4–24.7 17.6 (15.9) 20 10.4–24.8
SOL discrepancy
Prefeedback 36.3 (32.8) 18 28.8–43.8 35.9 (30.7) 20 21.5–50.3
Postfeedback 13.1 (23.7) 18 7.4–18.8 15.8 (24.4) 20 4.9–26.7
Subjective TST
Prefeedback 413.3 (56.1) 20 387.7–438.8 369.8 (56.9) 20 343.9–395.1
Postfeedback 422.2 (59.1) 20 389.8–454.5 408.6 (82.0) 20 376.2–440.9
Objective TST
Prefeedback 459.1 (63.4) 18 425.6–492.4 443.0 (74.8) 20 411.4–474.6
Postfeedback 423.3 (64.1) 18 390.4–456.1 442.8 (72.6) 20 411.6–473.9
TST discrepancy
Prefeedback 45.8 (50.0) 18 59.3–32.3 73.2 (62.5) 20 86.0–60.4
Postfeedback 1.1 (41.5) 18 13.3–11.1 34.2 (50.0) 20 57.7–10.6
Postexperiment ratings
Enjoyable?
c
7.5 (1.5) 21 7.2–7.9 6.5 (2.1) 21 6.1–7.0
Beneficial?
d
7.4 (1.9) 21 7.0–7.8 5.3 (2.8) 21 4.7–5.9
Acceptable to self?
e
7.6 (1.6) 21 7.1–7.9 6.5 (3.0) 21 5.8–7.1
Acceptable to others?
f
7.8 (1.6) 21 7.5–8.2 6.3 (3.0) 21 5.6–6.9
Note. Subjective sleep onset latency and total sleep time were reported to the nearest minute by the participants. CI confidence interval; APSQ
Anxiety and Preoccupation About Sleep Questionnaire; ISQ Insomnia Symptom Questionnaire; ISI Insomnia Severity Index; SOL sleep onset
latency (time from initial lights out to sleep onset); TST total sleep time (total amount of sleep obtained).
a
Response scale: 0 not at all well, 10 very well.
b
Response scale: 0 not at all, 10 very much.
c
Response scale: 0 not at all enjoyable,
10 very enjoyable.
d
Response scale: 0 I benefited nothing at all, 10 I benefited a lot.
e
Response scale: 0 not at all acceptable, 10 very
acceptable.
f
Response scale: 0 not at all acceptable, 10 very acceptable.
772
TANG AND HARVEY
Page 6
feedback group did not ( p .21). The effect size was 1.03 for the
behavioral experiment group and 0.28 for the verbal feedback
group.
Effect of the Feedback on the Subjective and Objective
Estimates of Sleep
Table 3 presents the mean subjective and objective estimates of
SOL and TST by group and session. No significant effects were
observed for these sleep measures except for a significant session
effect for the subjective estimates of SOL, F(1, 38) 8.3, p .01,
and TST, F(1, 38) 4.8, p .05, such that the subjective SOL
was estimated to be shorter and the subjective TST was estimated
to be longer following the feedback session.
Effect of the Feedback on the Discrepancy Between
Subjective and Objective Sleep Estimates
To index the misperception of SOL and TST, we calculated a
discrepancy score by subtracting the objective from the subjective
sleep estimate for SOL and TST. A positive value denotes an
overestimation, whereas a negative value denotes an underestima-
tion. The discrepancy scores for SOL and TST perception before
and after the feedback session are also presented in Table 3. The
analysis was based on the sleep data of 18 participants in the
behavioral experiment group and 20 participants in the verbal
feedback group. Missing data (n 4) were due to actigraphic file
corruption and missing sleep diary entries.
For the discrepancy between subjective and objective SOL
estimates, no significant main effect for group was noted. There
was a significant main effect for session, F(1, 36) 13.5, p .01,
such that the discrepancy was smaller after the feedback session
compared with before. The interaction effect was not significant.
For the discrepancy between subjective and objective TST esti-
mates, no significant main effect for group was noted. There was
a significant main effect for session, F(1, 36) 20.5, p .001,
such that the discrepancy was smaller after the feedback session
compared with before. The interaction effect was not significant.
Postexperiment Ratings
The mean values for the four postexperiment ratings are also
displayed in Table 3. Although the behavioral experiment group
gave higher ratings on each of these scales, independent-samples t
tests indicated that there were significant between-groups differ-
ences for the “beneficial,” t(40) 2.8, p .01, and “acceptable to
others,” t(30.3) 2.1, p .05, ratings. The differences for the
“enjoyable,” t(40) 1.8, p .08, and “acceptable to self,”
t(30.8) 1.5, p .14, ratings did not reach statistical significance.
Discussion
The first hypothesis tested was that, relative to those who were
simply told about the discrepancy, the participants who were
shown the discrepancy through a behavioral experiment would
think and feel more positively about their sleep. Consistent with
this hypothesis, the behavioral experiment group showed a greater
improvement in their sleep-perception rating and greater decrease
in their APSQ, ISQ, ISI, and sleep-distress rating than the verbal
feedback group. The magnitude of the improvement on these
self-report measures observed in the behavioral experiment group
(effect sizes: 0.79 –1.25) was large (Cohen, 1992) and was consis-
tently larger compared with the verbal feedback group (effect
sizes: 0.06 0.31). These results suggest that the behavioral
experiment is more potent than verbal techniques in improving
sleep perception and reducing sleep-related anxiety and distress in
patients with insomnia. Taken together, the findings corroborate
previous research showing that correcting sleep misperception can
bring about therapeutic benefits (Downey & Bonnet, 1992; Tang &
Harvey, 2004a) and provide empirical support to the proposal that
behavioral experiments can be a highly effective method, more
effective than verbal techniques, for promoting therapeutic change
(e.g., J. S. A. Beck, 1995).
The second hypothesis tested was that the participants who were
shown the discrepancy through a behavioral experiment would
estimate their sleep more accurately than those who were simply
told of the discrepancy. Contrary to this hypothesis, all participants
showed improvement in their subjective sleep estimates following
the feedback session, and both methods of feedback were equally
successful in minimizing the participants’ misperception of SOL
and TST. There are at least two possible explanations for these
findings. First, consistent with many previous studies (see Cham-
bers & Keller, 1993, for a review), as a group, our participants did
not exhibit a substantial sleep deficit at baseline according to the
actigraphic sleep estimates. Hence, there may have been a floor
effect on the improvement that can be shown in the sleep variables
postfeedback. Second, the observation period was only 2 days, so
it is plausible that a differential effect between the behavioral
experiment and verbal feedback did not emerge until later. How-
ever, it is just as possible that the findings reflected the reality—
that is, the verbal feedback was as effective as the behavioral
experiment in altering sleep misperception. If so, clinically, this is
good news, because the present results indicate that simple verbal
feedback of the discrepancy between subjective and objective
sleep estimates could reduce sleep misperception, even though it
might not be able to instigate significant changes in the way
patients think and feel about their sleep, as could a well-
implemented behavioral experiment.
Treatment acceptability is an issue that is often overlooked in
outcome research, although it is a criterion that has been used to
evaluate and compare strengths of different treatment strategies for
insomnia (Morin, Gaulier, Barry, & Kowatch, 1992). In the present
study, postexperiment ratings revealed that from the participant’s
perspective, the behavioral experiment was a treatment strategy
more “beneficial” and “acceptable” than verbal feedback. These
findings lend support to the advice of leading clinicians who
advocate the use of behavioral experiments as highly acceptable
treatment strategies (e.g., J. S. A. Beck, 1995; Greenberger &
Padesky, 1995).
Several limitations of the current study deserve comment. First,
we used a relatively small insomnia sample recruited from a
predominantly White, middle-class community. Hence, the find-
ings must be interpreted with caution as to their generalizability.
Although the current sample was a step closer to the target pop-
ulation compared with the sample used in a previous study (Tang
& Harvey, 2004a), it will be important to confirm the generaliz-
ability of our results to a treatment-seeking sample. Having said
that, we note that the participants included in this study are
773
BEHAVIORAL EXPERIMENT VERSUS VERBAL FEEDBACK
Page 7
comparable to treatment seekers on several critical dimensions:
meeting strict DSM–IV–TR criteria for insomnia, fulfilling the
World Health Organization frequency and duration specifications,
scoring higher than 5 on the PSQI, and sharing characteristics (as
indicated by BAI, BDI, and PSWQ scores) possessed by
treatment-seeking patients with insomnia (e.g., Espie, Inglis,
Tessier, & Harvey, 2001). Nonetheless, to check that the current
findings can be applied to a more heterogeneous group, it will be
essential to replicate the current study in a larger, more ethnically
and socially diverse treatment-seeking sample.
Second, although the participants were not told that the true aim
of the experiment was to alter their sleep perception, we cannot
rule out a possible experimental expectation effect—that is, the
participants might have shown improvement because they were
expected to. Although several active counterdemand measures
were taken (e.g., using a cover story and a standardized procedure
with limited interaction between the experimenter and the partic-
ipants), future research should consider taking even more rigorous
checks to ensure that the therapeutic effects were not inflated by
the participants’ expectations.
Third, we used actigraphy in this study to provide nonintrusive
estimates of SOL and TST. As we discussed in the Materials
section, actigraphy is sensitive in detecting sleep but less specific
in detecting wake and sleep onset identified by PSG. Actigraphy
may thus be less reliable in people who can lie immobile for an
extended period of time (e.g., Sadeh & Acebo, 2002). As such, the
subjective– objective sleep estimates discrepancy adopted in this
study as an index of sleep misperception should be interpreted
within the confines of the limitations of actigraphy. Future re-
search using more precise sleep estimates (e.g., PSG) should be
conducted. It is necessary to point out, however, that all sleep-
estimating technology has its advantages and disadvantages; even
PSG, the current gold standard, suffers from problems such as high
intrusiveness, considerable inconsistency in the way PSG is ad-
ministered and scored, and heavy capital and overhead costs.
Actigraphy was used in this study because it is nonintrusive and
relatively less expensive. Most important, the way the actigraphic
data are analyzed and presented fit the format and goal of the
behavioral experiment.
Future research in this domain is likely to be fruitful. Four
specific suggestions are offered. First, in the present study, to
match with the length and the format of the behavioral experiment
and to control for nonspecific confounding factors, the verbal
feedback group were told about the discrepancy after 25 min of
waiting. It is noteworthy that this procedure does not reflect the
procedure that would be adopted in clinical practice and may have
affected the acceptability of the intervention. Similarly, the proce-
dure adopted for the behavioral experiment group was, strictly
speaking, an analogue of a clinical behavioral experiment, as the
participants were unaware that the purpose of the study was to test
the accuracy of their cognition about sleep. Therefore, it will be
important to replicate the present findings with a behavioral ex-
periment delivered in a clinical format, with the patients knowing
the rationale behind the behavioral experiment. Second, the dura-
tion of the postfeedback period was only 2 days, and the durability
of the therapeutic effects is not known, so longer term follow-up to
test the stability of the outcome is required. To gauge the extent of
clinical improvement, future studies should include an insomnia
status assessment at posttreatment and at follow-up. Third, as
misperception of sleep occurs along a continuum (Edinger & Fins,
1995) 6 participants were excluded from the final analyses for not
displaying the typical pattern of sleep misperception. It is likely
that not all insomnia patients who present for treatment do over-
estimate their SOL and underestimate their TST. Accordingly, we
recommend that an assessment for the presence of misperception
be conducted before this behavioral experiment is administered.
Finally, and related to the last point, sleep misperception in in-
somnia is likely to be a complex phenomenon. We know from
previous research that sleep misperception does not occur exclu-
sively in patients diagnosed with sleep state misperception (Car-
skadon et al., 1976; Edinger & Fins, 1995). There is also a
promising line of research (e.g., Perlis et al., 2001) that raises the
possibility that the phenomenon commonly referred to as sleep
misperception may be accounted for by qualitative aspects of sleep
that are not captured in conventional scoring.
In summary, although both techniques effectively reduced mis-
perception of sleep, the behavioral experiment was superior to
verbal feedback because it also changed how participants thought
and felt about their sleep. Specifically, the behavioral experiment
was associated with greater reduction in the levels of sleep im-
pairment, insomnia symptoms, and sleep-related anxiety and dis-
tress, and it was potent in helping the patients to acquire a more
positive perception of their sleep. Theoretically, these findings
lend empirical support to the notion that behavioral experiments
are more powerful, relative to verbal techniques, for instigating
positive cognitive and emotional changes (e.g., J. S. A. Beck,
1995; Clark, 1999; Greenberger & Padesky, 1995). Clinically, the
findings suggest that a behavioral experiment targeting mispercep-
tion of sleep may be effective in markedly reducing the subjective
concern and distress of individuals with insomnia and in improv-
ing their subjective estimates of sleep, even though no immediate
changes were observed for objective sleep estimates. An important
next step is to examine whether this novel behavioral experiment
might improve upon other known efficacious insomnia treatments
for individuals diagnosed with sleep state misperception and
whether it might enhance the already moderately effective
cognitive– behavioral intervention package for insomnia (average
effect size 0.42–0.88; Morin et al., 1994). In particular, it is of
clinical interest to check in future research whether the reduction
in sleep-related anxiety and distress associated with the adminis-
tration of this behavioral experiment would (a) have a positive
effect on objective sleep measures in the longer term and (b) help
more individuals treated for insomnia to acquire “good sleeper”
status. It is also noted that in the absence of resources to provide
objective feedback of the discrepancy, the present results indicate
that some improvement occurs with simple verbal feedback. Given
the encouraging results generated by the present study, future
research should be directed toward the evaluation of other behav-
ioral experiments that may be useful to treat patients with insomnia
(see Ree & Harvey, 2004).
References
American Psychiatric Association. (2000). Diagnostic and statistical man-
ual of mental disorders DSM–IV–TR (4th ed., text rev.). Washington,
DC: Author.
Ancoli-Israel, S., Clopton, P., Klauber, M. R., Fell, R., & Mason, W.
(1997). Use of wrist activity for monitoring sleep/wake in demented
nursing-home patients. Sleep, 20, 24 –27.
774
TANG AND HARVEY
Page 8
Bastien, C. H., Vallieres, A., & Morin, C. M. (2001). Validation of the
Insomnia Severity Index as an outcome measure for insomnia research.
Sleep Medicine, 2, 297–307.
Beck, A. T., Epstein, N., Brown, G., & Steer, R. A. (1988). An inventory
for measuring clinical anxiety: Psychometric properties. Journal of
Consulting and Clinical Psychology, 56, 893–897.
Beck, A. T., Rush, A. J., Shaw, B. F., & Emery, G. (1979). Cognitive
therapy of depression. 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.
Beck, J. S. A. (1995). Cognitive therapy: Basics and beyond. New York:
Guilford Press.
Bennett-Levy, J., Butler, G., Fennell, M., Hackmann, A., Mueller, M., &
Westbrook, D. (Eds.). (2004). Oxford guide to behavioral experiments in
cognitive therapy. Oxford, England: Oxford University Press.
Bonnet, M. H., & Arand, D. L. (1994). Impact of the level of physiological
arousal on estimates of sleep latency. In R. D. Ogilvie & J. R. Harsh
(Eds.), Sleep onset: Normal and abnormal processes (pp. 127–139).
Washington, DC: American Psychological Association.
Bonnet, M. H., & Arand, D. L. (1996). The consequences of a week of
insomnia. Sleep, 19, 453– 461.
Borkovec, T. D. (1982). Insomnia. Journal of Consulting and Clinical
Psychology, 50, 880 895.
Buysse, D. J., Reynolds, C. F., III, Monk, T. H., Berman, S. R., & Kupfer,
D. J. (1989). The Pittsburgh Sleep Quality Index: A new instrument for
psychiatric practice and research. Psychiatry Research, 28, 193–213.
Carpenter, J. S., & Andrykowski, M. A. (1998). Psychometric evaluation
of the Pittsburgh Sleep Quality Index. Journal of Psychosomatic Re-
search, 45, 5–13.
Carskadon, M. A., Dement, W. C., Mitler, M. M., Guilleminault, C.,
Zarcone, V. P., & Spiegel, R. (1976). Self-reports versus sleep labora-
tory findings in 122 drug-free subjects with complaints of chronic
insomnia. American Journal of Psychiatry, 133, 1382–1388.
Chambers, M. J., & Keller, B. (1993). Alert insomniacs: Are they really
sleep deprived? Clinical Psychology Review, 13, 667–682.
Clark, D. M. (1999). Anxiety disorders: Why they persist and how to treat
them. Behavior Research and Therapy, 37(Suppl. 1), S5–S27.
Cohen, J. (1992). Quantitative methods in psychology: A power primer.
Psychological Bulletin, 112, 155–159.
Cole, R. J., Kripke, D. F., Gruen, W., Mullaney, D. J., & Gillin, J. C.
(1992). Automatic sleep/wake identification from wrist activity. Sleep,
15, 461–469.
DeSouza, L., Benedito-Silva, A. A., Pires, M. L., Poyares, D., Tufik, S., &
Calil, H. M. (2003). Further validation of actigraphy for sleep studies.
Sleep, 26, 81– 85.
Downey, R., III, & Bonnet, M. H. (1992). Training subjective insomniacs
to accurately perceive sleep onset. Sleep, 15, 58 63.
Edinger, J. D., & Fins, A. I. (1995). The distribution and clinical signifi-
cance of sleep time misperceptions among insomniacs. Sleep, 18, 232–
239.
Espie, C. A., Inglis, S. J., Tessier, S., & Harvey, L. (2001). The clinical
effectiveness of cognitive behavior therapy for chronic insomnia: Im-
plementation and evaluation of a sleep clinic in general medical practice.
Behaviour Research and Therapy, 39, 45–60.
Ford, D. E., & Kamerow, D. B. (1989). Epidemiologic study of sleep
disturbances and psychiatric disorders: An opportunity for prevention?
JAMA, 262, 1479 –1484.
Frankel, B. L., Buchbinder, R., Coursey, R., & Snyder, F. (1976). Recorded
and reported sleep in chronic primary insomnia. Archives of General
Psychiatry, 33, 615– 623.
Greenberger, D., & Padesky, C. A. (1995). Mind over mood: Change how
you feel by changing the way you think. New York: Guilford Press.
Harvey, A. G. (2002). A cognitive model of insomnia. Behaviour Research
and Therapy, 40, 869 893.
Harvey, A. G., Clark, D. M., Ehlers, A., & Rapee, R. M. (2000). Social
anxiety and self-impression: Cognitive preparation enhances the bene-
ficial effects of video feedback following a stressful social task. Behav-
iour Research and Therapy, 38, 1183–1192.
Harvey, A. G., Nelson, J., Neitzert Semler, C., Tang, N. K. Y., Sharpley,
A., & Ree, M. (n.d.). Validation of the Insomnia Diagnostic Interview.
Unpublished manuscript.
Harvey, A. G., & Tang, N. K. Y. (2003). Cognitive behaviour therapy for
insomnia: Can we rest yet? Sleep Medicine Reviews, 7, 237–262.
Kuishida, C. A., Chang, A., Gadkary, C., Guilleminault, C., Carrillo, O., &
Dement, W. C. (2001). Comparison of actigraphic, polysomnographic,
and subjective assessment of sleep parameters in sleep-disordered pa-
tients. Sleep Medicine, 2, 389 –396.
Lichstein, K. L., Durrence, H. H., Taylor, D. J., Bush, A. J., & Riedel,
B. W. (2003). Quantitative criteria for insomnia. Behaviour Research
and Therapy, 41, 427– 445.
Lundh, L. G., & Broman, J. E. (2000). Insomnia as an interaction between
sleep-interfering and sleep-interpreting processes. Journal of Psychoso-
matic Research, 49, 299 –310.
Meyer, T. J., Miller, M. L., Metzger, R. L., & Borkovec, T. D. (1990).
Development and validation of the Penn State Worry Questionnaire.
Behaviour Research and Therapy, 28, 487–495.
Morin, C. M. (1993). Insomnia: Psychological assessment and manage-
ment. New York: Guilford Press.
Morin, C. M., Culbert, J. P., & Schwartz, S. M. (1994). Nonpharmaco-
logical interventions for insomnia: A meta-analysis of treatment effi-
cacy. American Journal of Psychiatry, 151, 1172–1180.
Morin, C. M., Gaulier, B., Barry, T., & Kowatch, R. A. (1992). Patients’
acceptance of psychological and pharmacological therapies for insom-
nia. Sleep, 15, 302–305.
Morin, C. M., Hauri, P. J., Espie, C. A., Spielman, A. J., Buysse, D. J., &
Bootzin, R. R. (1999). Nonpharmacologic treatment of chronic insom-
nia. Sleep, 22, 1134 –1156.
Murtagh, D. R., & Greenwood, K. M. (1995). Identifying effective psy-
chological treatments for insomnia: A meta-analysis. Journal of Con-
sulting and Clinical Psychology, 63, 7989.
Neitzert-Semler, C., & Harvey, A. G. (2005). Misperception of sleep can
adversely affect daytime functioning in insomnia. Behaviour Research
and Therapy, 43, 843– 856.
Ohayon, M. M., Caulet, M., Priest, R. G., & Guilleminault, C. (1997).
DSM–IV and ICSD–90 insomnia symptoms and sleep dissatisfaction.
British Journal of Psychiatry, 171, 382–388.
Perlis, M. L., Giles, D. E., Mendelson, W. B., Bootzin, R. R., & Wyatt,
J. K. (1997). Psychophysiological insomnia: The behavioral model and
a neurocognitive perspective. Journal of Sleep Research, 6, 179 –188.
Perlis, M. L., Smith, M. T., Andrews, P. J., Orff, H., & Giles, D. E. (2001).
Beta/gamma EEG activity in patients with primary and secondary in-
somnia and good sleeper controls. Sleep, 24, 110 –117.
Ree, M., & Harvey, A. G. (2004). Insomnia. In J. Bennett-Levy, G. Butler,
M. Fennell, A. Hackmann, M. Mueller, & D. Westbrook (Eds.), Oxford
guide to behavioral experiments in cognitive therapy (pp. 287–305).
Oxford, England: Oxford University Press.
Reid, K., & Dawson, D. (1999). Correlation between wrist activity monitor
and electrophysiological measures of sleep in a simulated shiftwork
environment for younger and older subjects. Sleep, 22, 378 –385.
Roth, T., & Ancoli-Israel, S. (1999). Daytime consequences and correlates
of insomnia in the United States: Results of the 1991 National Sleep
Foundation Survey II. Sleep, 22, S354–S358.
Sadeh, A., & Acebo, C. (2002). The role of actigraphy in sleep medicine.
Sleep Medicine Reviews, 6, 113–124.
Sadeh, A., Hauri, P. J., Kripke, D. F., & Lavie, P. (1995). The role of
actigraphy in the evaluation of sleep disorders. Sleep, 18, 288 –302.
775
BEHAVIORAL EXPERIMENT VERSUS VERBAL FEEDBACK
Page 9
Sadeh, A., Sharkey, K. M., & Carskadon, M. A. (1994). Activity-based
sleep–wake identification: An empirical test of methodological issues.
Sleep, 17, 201–207.
Salkovskis, P. M. (1991). The importance of behavior in the maintenance
of anxiety and panic: A cognitive account. Behavioral Psychotherapy,
19, 6–19.
Salkovskis, P. M., Clark, D. M., Hackmann, A., Wells, A., & Gelder, M. G.
(1999). An experimental investigation of the role of safety-seeking
behaviors in the maintenance of panic disorder with agoraphobia. Be-
haviour Research and Therapy, 37, 559–574.
Simon, G. E., & Von Korff, M. (1997). Prevalence, burden and treatment
of insomnia in primary care. American Journal of Psychiatry, 154,
1417–1423.
Smith, S., & Trinder, J. (2000). The effect of arousals during sleep onset on
estimates of sleep onset latency. Journal of Sleep Research, 9, 129 –135.
Spielman, A. J., Saskin, P., & Thorpy, M. J. (1987). Treatment of chronic
insomnia by restriction of time in bed. Sleep, 10, 45–56.
Stoller, M. K. (1994). Economic effects of insomnia. Clinical Therapeu-
tics: The International Peer Reviewed Journal of Drug Therapy, 16,
873– 897.
Tang, N. K. Y., & Harvey, A. G. (2004a). Correcting distorted perception
of sleep in insomnia: A novel behavioral experiment? Behaviour Re-
search and Therapy, 42, 27–39.
Tang, N. K. Y., & Harvey, A. G. (2004b). Effects of cognitive arousal and
physiological arousal on sleep perception. Sleep, 27, 69 –78.
Van Egeren, L., Haynes, S. N., Franzen, M., & Hamilton, J. (1983).
Presleep cognitions and attributions in sleep-onset insomnia. Journal of
Behavioral Medicine, 6, 217–232.
Webster, J. B., Kripke, D. F., Messin, S., Mullaney, D. J., & Wyborney, G.
(1982). An activity-based sleep monitor system for ambulatory use.
Sleep, 5, 389 –399.
World Health Organization. (1992). The ICD–10 classification of mental
and behavioral disorders: Diagnostic criteria for research (10th rev.).
Washington, DC: Author.
Received November 29, 2004
Revision received February 24, 2006
Accepted March 6, 2006
776
TANG AND HARVEY
Page 10
  • Source
    • "Výzkumy také ukazují na rozdíl mezi subjektivním a objektivním hodnocením spánku. Mnoho pacientù s nespavostí nadhodnocuje dobu do usnutí a podhodnocuje celkovou dobu spánku ve srovnání s výsledky objektivního mìøení pomocí polysomnografie – PSG (Schneider-Helmert a Kumar, 1995; Morgenthaler, 2006) – i aktigrafie (Vanable et al., 2000; Tang a Harvey, 2006; Van den Berg, 2008 ). Tato diskrepance se oznaèuje jako spánková mispercepce a objevuje se u rùzných typù insomnie. "
    [Show abstract] [Hide abstract] ABSTRACT: Insomnia is one of the most common sleep disorders. The diagnosis is based on patient's subjective statement. The objective findings of sleep architecture in insomnia are not uniform. These patients often show the discrepancy between subjective and objective evaluation of sleep. The cause of the above-described sleep state misperception is not fully known. It is assumed that both psychological and somatic factors play an important role in its development. The aim of this article is to summarize current knowledge and to contribute to the understanding of this phenomenon. Studying the difference between subjective and objective evaluation of sleep has clinical, theoretical and public health importance. Recent studies have shown that sleep state misperception may be a risk factor in the development of anxiety and insomnia with serious objective sleep deficit. Understanding the causes may improve the treatment of one of the common health problems.
    Full-text · Article · Aug 2015
  • Source
    • "This study supports the hypothesis that sleep discrepancy is temporally related to sleep quality. Some evidence suggests that training participants to report subjective estimates of SOL and WASO to match objective measures may have therapeutic effects (Downey and Bonnet, 1992; Mercer et al., 2002; Tang and Harvey, 2006). These studies did not include a follow-up assessment to determine long-term effects. "
    [Show abstract] [Hide abstract] ABSTRACT: Discrepancy between subjective and objective measures of sleep is associated with insomnia and increasing age. Cognitive behavioural therapy for insomnia improves sleep quality and decreases subjective–objective sleep discrepancy. This study describes differences between older adults with insomnia and controls in sleep discrepancy, and tests the hypothesis that reduced sleep discrepancy following cognitive behavioural therapy for insomnia correlates with the magnitude of symptom improvement reported by older adults with insomnia. Participants were 63 adults >60 years of age with insomnia, and 51 controls. At baseline, participants completed sleep diaries for 7 days while wearing wrist actigraphs. After receiving cognitive behavioural therapy for insomnia, insomnia patients repeated this sleep assessment. Sleep discrepancy variables were calculated by subtracting actigraphic sleep onset latency and wake after sleep onset from respective self-reported estimates, pre- and post-treatment. Mean level and night-to-night variability in sleep discrepancy were investigated. Baseline sleep discrepancies were compared between groups. Pre–post-treatment changes in Insomnia Severity Index score and sleep discrepancy variables were investigated within older adults with insomnia. Sleep discrepancy was significantly greater and more variable across nights in older adults with insomnia than controls, P ≤ 0.001 for all. Treatment with cognitive behavioural therapy for insomnia was associated with significant reduction in the Insomnia Severity Index score that correlated with changes in mean level and night-to-night variability in wake after sleep onset discrepancy, P < 0.001 for all. Study of sleep discrepancy patterns may guide more targeted treatments for late-life insomnia.
    Full-text · Article · Sep 2014 · Journal of Sleep Research
  • Source
    • "Bipolar patients may be prone to misperceive sleep for wakefulness [3,32]. There will also be a focus on challenging and correcting sleep state misperception using cognitive and behavioral techniques [33]. Circadian rhythm is likely to be disturbed in the included patients, so there will be particular emphasis on stabilization of the circadian rhythm. "
    [Show abstract] [Hide abstract] ABSTRACT: Patients with bipolar disorder experience sleep disturbance, even in euthymic phases. Changes in sleep pattern are frequent signs of a new episode of (hypo)mania or depression. Cognitive behavioral therapy for insomnia (CBT-I) is an effective treatment for primary insomnia, but there are no published results on the effects of CBT-I in patients with bipolar disorder. In this randomized controlled trial, we wish to compare CBT-I and treatment as usual with treatment as usual alone to determine its effect in improving quality of sleep, stabilizing minor mood variations and preventing new mood episodes in euthymic patients with bipolar disorder and comorbid insomnia. Patients with euthymic bipolar I or II disorder and insomnia, as verified by the Structured Clinical Interview for DSM Disorders (SCID-1) assessment, will be included. The patients enter a three-week run-in phase in which they complete a sleep diary and a mood diary, are monitored for seven consecutive days with an actigraph and on two of these nights with polysomnography in addition before randomization to an eight-week treatment trial. Treatment as usual consists of pharmacological and supportive psychosocial treatment. In this trial, CBT-I will consist of sleep restriction, psychoeducation about sleep, stabilization of the circadian rhythm, and challenging and correcting sleep state misperception, in three to eight sessions. This trial could document a new treatment for insomnia in bipolar disorder with possible effects on sleep and on stability of mood. In addition, more precise information can be obtained about the character of sleep disturbance in bipolar disorder.Trial registration: ClinicalTrials.gov: NCT01704352.
    Full-text · Article · Jan 2014 · Trials
Show more