Testing a Model of Pain Appraisal and Coping in Children With Chronic
Lynn S. Walker
Vanderbilt University School of Medicine
Craig A. Smith, Judy Garber, and
Robyn Lewis Claar
This prospective study of children with recurrent abdominal pain (N ? 133; ages 8–15 years) used path
analysis to examine relations among dispositional pain beliefs and coping styles, cognitions and behavior
related to a specific pain episode, and short- and long-term outcomes. Children believing they could not
reduce or accept pain appraised their episode-specific coping ability as low and reported passive coping
behavior. Dispositional passive coping had direct effects on both episode-specific passive coping and
long-term symptoms and disability. Accommodative coping (acceptance and self-encouragement) was
associated with reduced episode-specific distress, which itself predicted reduced depressive symptoms 3
months later. Results suggest that coping-skill interventions for children with chronic pain should target
reductions in passive coping and consider the potential benefits of accommodative coping strategies.
Keywords: pain, pain beliefs, appraisal, coping, children
Studies of children’s coping with chronic pain typically have
conceptualized coping at the dispositional level and have exam-
ined the relation of coping styles to broad outcomes such as
symptom severity, disability, and health service utilization (e.g.,
Gil, Williams, Thompson, & Kinney, 1991; Thomsen et al., 2002;
Walker, Smith, Garber, & Van Slyke, 1997). The process by which
children’s styles of coping with pain may lead to these outcomes
rarely has been examined. However, it is reasonable to assume that
children’s coping styles predict their episode-specific coping strat-
egies and these, in turn, determine whether the outcomes of indi-
vidual pain episodes will be positive or negative, in a process that
repeats itself multiple times to yield long-term outcomes. For
example, children who report on a retrospective questionnaire that
they usually go to bed when they experience pain would be
expected to report in a diary assessment that they had gone to bed
on the day of a pain episode. Of course, pain episodes may vary
considerably across time and circumstances, and this would be
reflected in variability in the coping strategies a child might use
during a particular pain episode. Thus, the correspondence be-
tween dispositional and episode-specific coping would never be
The goal of this study was to test a conceptual model of the
relations among dispositional pain beliefs and coping styles, cog-
nitions and behavior related to a specific pain episode, and short-
and long-term outcomes. The relations in this model are important
on a conceptual level for understanding the process by which
appraisal and coping with pain influence children’s health and on
a practical level for evaluating the use of children’s retrospective
reports of their usual coping style in predicting their behavior
during specific episodes of pain. The relation of dispositional
styles of coping to episode-specific coping has been examined in
the general stress and coping literature (e.g., Carver & Scheier,
1994; Schwartz, Neale, Marco, Shiffman, & Stone, 1999) and is
the focus of continued attention in a debate regarding the consis-
tency of coping behavior and whether it can be regarded as a trait
(e.g., Ptacek & Gross, 1997; Schwartz et al., 1999). Little is
known, however, about the relation between dispositional styles of
coping with pain and coping associated with individual pain epi-
sodes (Rudolph, Dennig, & Weisz, 1995).
General Conceptual Framework
The proposed model is based on the appraisal and coping
framework advanced by Lazarus and Folkman (1984; Folkman,
Lazarus, Dunkel-Schetter, DeLongis, & Gruen, 1986). This frame-
work emphasizes the importance of the individual’s perceptions in
evaluating potential stressors (cf. Monroe & Kelley, 1995). Laza-
rus and Folkman conceptualized subjective evaluation of stressors
as appraisal of coping potential, that is, the individual’s evaluation
of “which coping options are available, the likelihood that a given
coping option will accomplish what it is supposed to, and the
likelihood that one can apply a particular strategy or set of strat-
egies effectively” (p. 35). In the context of pain, appraisal of
coping potential has been investigated elsewhere as pain self-
efficacy (Lefebvre et al., 1999). According to the framework put
forward by Lazarus and Folkman, appraisals of coping potential
predict the nature of coping strategies individuals use in confront-
ing a particular stressor. These strategies, in turn, predict
Lynn S. Walker, Division of Adolescent Medicine and Behavioral
Science, Department of Pediatrics, Vanderbilt University School of Med-
icine; Craig A. Smith, Judy Garber, and Robyn Lewis Claar, Department of
Psychology and Human Development, Vanderbilt University.
Robyn Lewis Claar is now at the Department of Psychiatry, Harvard
University Medical School and Children’s Hospital, Boston.
This research was supported by National Institute on Child Health and
Human Development Grants R01 HD23264 and P30 HD15052.
Correspondence concerning this article should be addressed to Lynn S.
Walker, Division of Adolescent Medicine and Behavioral Science, Depart-
ment of Pediatrics, Vanderbilt University School of Medicine, Nashville,
TN 37232-3571. E-mail: firstname.lastname@example.org
2005, Vol. 24, No. 4, 364–374
Copyright 2005 by the American Psychological Association
0278-6133/05/$12.00 DOI: 10.1037/0278-6126.96.36.1994
This conceptual framework can be applied both at the disposi-
tional level in understanding how coping beliefs and styles of
coping with chronic stressors influence global outcomes and at the
situational level in understanding how cognitions and behaviors
associated with a particular stressor episode influence the imme-
diate outcome of that episode. We have combined both levels of
analysis in a model in which chronic pain is the stressor. Prior
pediatric investigations have examined the relation of pain coping
styles to adjustment (e.g., Gil et al., 1991; Thomsen et al., 2002;
Walker et al., 1997) and the relation of pain cognitions to distress
during a single episode of pain (Claar, Walker, & Smith, 2002),
but this is the first to examine both appraisal and coping with pain
and to include both dispositional and situational levels of analysis
in the same model.
Our general conceptual model is depicted in Figure 1. Pain
beliefs and coping styles are hypothesized to influence appraisals
and coping strategies associated with a single pain episode.
Episode-specific appraisals should further influence the specific
coping strategies used during the pain episode, and these coping
activities, in turn, are hypothesized to influence acute outcomes of
the pain episode. Finally, episode-specific outcomes are hypothe-
sized to contribute to long-term outcomes including somatic symp-
toms, functional disability, and depressive symptoms. Although
the figure depicts a single pain episode, it is intended to represent
a process that is repeated multiple times in chronic or recurrent
Of course, the relations of dispositional pain beliefs and coping
styles to episode-specific appraisals and coping behaviors would
be attenuated by contextual factors that influence responses unique
to the particular episode. Similarly, we reasoned that only in
extreme circumstances would the outcomes of a single encounter
strongly influence long-term outcomes. Instead, it is the pattern of
outcomes over a series of episodes that is hypothesized to shape
long-term outcomes. Thus, we expected the observed relations
between episode-specific outcomes and long-term outcomes to be
weak, as represented in Figure 1 by dashed lines from episode-
specific outcomes to long-term outcomes.
A Combined Dispositional–Situational Model
We translated the general theoretical framework into a testable
model, with each construct in Figure 1 represented by measurable
variables. We tested this model in a sample of children with
recurrent abdominal pain (RAP), the most common recurrent pain
condition of childhood (Apley, 1975; McGrath, 1994). The study
builds on prior investigations demonstrating that dispositional cop-
ing styles of children with RAP predict long-term health outcomes
(Walker et al., 1997). No prior investigations have examined how
children with RAP cope with individual pain episodes.
Appraisals of coping potential were assessed as the child’s
perceived ability to engage in the two major forms of coping
described by Lazarus and Folkman (1984). Thus, appraisal of
problem-focused coping potential (PFCP) refers to the ability to
alter circumstances to make them more desirable (in this case, to
alleviate pain), whereas appraisal of emotion-focused coping po-
tential (EFCP) refers to the ability to accept and adjust to circum-
stances, even if they cannot be improved (in this case, to accept
and adjust to pain). Prior literature has only examined global
self-efficacy beliefs regarding pain coping (Jensen, Turner, &
Romano, 1991; Lefebvre et al., 1999; Thompson, Gil, Abrams, &
long-term outcomes were expected to be weak.
Conceptual model. The dashed lines indicate that the observed relations between short- and
TESTING A MODEL OF PAIN APPRAISAL AND COPING
Phillips, 1992). However, the distinction between PFCP and EFCP
is important because current pain management approaches empha-
size that efforts to eradicate pain (problem-focused coping) may
not be as adaptive as accepting pain and maximizing one’s activ-
ities (i.e., emotion-focused, accommodative coping; e.g., Bursch,
Walco, & Zeltzer, 1998; McCracken, 1998).
Appraisals of coping potential were expected to predict three
broad categories of coping that we have previously applied to
pediatric pain (Walker et al., 1997). Children who believed that
they had the ability to reduce or eliminate their pain (i.e., high
PFCP) were hypothesized to engage in problem-focused, coping
strategies that pain coping literature has referred to as active
coping (e.g., Brown & Nicassio, 1987; Walker et al., 1997).
Children who believed that they could not reduce their pain (i.e.,
low PFCP) but could accept and adjust to pain (i.e., high EFCP)
were expected to engage in a range of emotion-focused strategies
that entailed accommodation to pain and which we have previ-
ously referred to as accommodative coping (Walker et al., 1997).
Finally, children who believed that they could neither reduce their
pain (i.e., low PFCP) nor accept and adjust to their pain (i.e., low
EFCP) were expected to respond to pain with coping strategies
characterized by passivity, negative cognitions, and lack of active
problem solving. This form of coping reflects a second type of
emotion-focused coping in the Lazarus and Folkman (1984)
framework that has been referred to as passive coping in the pain
literature (e.g., Brown & Nicassio, 1987; Walker et al., 1997). To
maintain the distinction between accommodative and passive
forms of emotion-focused coping, below we refer to the modes of
coping by the names used in the pain coping literature: active,
accommodative, and passive coping.
In the pain coping literature (e.g., Brown & Nicassio, 1987),
passive coping, which involves such strategies as taking to bed,
restricting one’s activities, and assuming the worst,1has been
associated with psychological distress and functional impairment
(e.g., Brown, Nicassio, & Wallston, 1989; Gil et al., 1991; Smith,
Wallston, Dwyer, & Dowdy, 1997; Turner, 1991). Active coping,
which involves problem-solving strategies aimed at reducing pain,
has been associated with more positive outcomes (e.g., Brown &
Nicassio, 1987; Gil et al., 1991; Smith et al., 1997). Finally,
accommodative coping, which theoretically enables one to adapt to
unchangeable stressors (see Lazarus, 1991; Smith & Lazarus,
1990), has been associated with beneficial outcomes (Turner,
1991; Walker et al., 1997).
All constructs were operationalized at the dispositional and
situational levels. Appraisals of coping potential were assessed at
the dispositional level as children’s pain coping beliefs and at the
situational level as children’s appraisals of their ability to cope
with a particular pain episode. Coping styles were assessed at the
dispositional level as the degree to which the child usually engaged
in each coping category when in pain. At the situational level,
coping was assessed in terms of coping strategies the child actually
used during a specific pain episode. Short-term outcomes were
represented by somatic symptoms and emotional distress that the
child reported experiencing during the pain episode. Long-term
outcomes were represented by measures of more chronic health
outcomes including somatic symptoms, functional disability, and
depressive symptoms that often are elevated in children with
persistent abdominal pain (Walker, Garber, & Greene, 1993).
Figure 2 summarizes the hypothesized relations among these
variables. We started with the simplifying assumption that each
dispositional variable would predict its situational counterpart but
only its counterpart.2Thus, for example, we predicted that a
passive coping style would be associated with passive coping
during a particular pain episode but not with active or accommo-
dative coping during the pain episode. Within the pain episode, we
expected episode-specific appraisals to be associated with episode-
specific coping. Specifically, we predicted that episode-specific
PFCP would be positively associated with episode-specific active
coping because active coping involves problem-solving efforts to
alleviate pain (Walker et al., 1997). Episode-specific EFCP was
expected to be positively associated with accommodative coping
as it involves emotional acceptance of pain. When appraisal of
PFCP is low, accommodative coping is more likely (Folkman,
1984; Lazarus & Folkman, 1984), and thus we also predicted that
PFCP would be negatively associated with accommodative coping.
Finally, we predicted that low levels of both PFCP and EFCP
would be associated with passive coping, as passive coping reflects
All three types of episode-specific coping efforts were expected
to predict short-term outcomes. Active coping and accommodative
coping were expected to have the greatest potential benefit and
thus to be negatively associated with somatic and affective symp-
toms. In line with prior literature, episode-specific passive coping
was expected to be associated with increased levels of both so-
matic and emotional distress. These short-term outcomes were
then predicted to be positively associated with long-term out-
comes. To simplify, we predicted that each short-term outcome
would be related to the long-term outcome(s) it most resembled.
For example, episode-specific somatic symptoms were predicted
to be positively associated with both long- term somatic symptoms
and functional disability but not with depressive symptoms.
1A reviewer noted that following Lazarus and Folkman (1984), to be
considered coping a strategy should be effortful, and the reviewer won-
dered whether some of the passive strategies such as assuming the worst
were sufficiently effortful to be considered coping. The point is arguable,
but we believe that the passive strategies we assessed do qualify as coping
as intended by Lazarus and Folkman. Effort was included as a criterion for
coping to distinguish coping from both automatized adjustive behaviors
and unconscious defense mechanisms (cf. Lazarus & Folkman, 1984). In
our view, cognitions such as assuming the worst are not unconscious and
need not be automatized. Furthermore, these cognitions are consistent with
existing conceptualizations of passive coping with pain (e.g., Brown &
Nicassio, 1987). Thus, we opted to risk being overly inclusive, rather than
overly exclusive, in our conceptualization of coping.
2This assumption and the other simplifying assumptions we describe
were made to keep the initial model we tested as simple and as straight-
forward as possible, while still doing justice to the more general model
depicted in Figure 1. In addition to considerations of parsimony, this was
done in light of the relatively small number of observations we had
available to test our model (see Method section). Our reasoning was that,
all else being equal, a simpler model would have a greater likelihood of
converging into a stable solution than a more complicated model. As
discussed in the analysis overview, sample-size considerations also led us
to model observed indicators instead of attempting to model latent
WALKER, SMITH, GARBER, AND CLAAR
The sample included 133 consecutive new patients who were referred to
a pediatric gastroenterology clinic for evaluation of abdominal pain. Pa-
tients were eligible if they had experienced chronic or recurrent episodes of
abdominal pain severe enough to interrupt activities and occurring over a
period of at least 3 months and if they were between 8 and 15 years of age,
the period when chronic or recurrent abdominal pain is most common (cf.
Apley, 1975). Exclusionary criteria included a chronic health condition or
mental retardation. Of the 229 patient families contacted, 57 (26%) did not
meet eligibility criteria and 18 (8%) declined, leaving 154 participants.
Complete data were obtained for 133 participants3who constituted the
final sample. The sample was primarily Caucasian (95%) and female
(57%), with a mean age of 10.80 years (SD ? 2.10).
Parents of children scheduled for evaluation of abdominal pain were
identified by clinic staff and contacted several days prior to their clinic
visit. Those who expressed interest in the study were screened for eligi-
bility and asked to arrive early if they wished to participate. Informed
consent was obtained at the clinic by research staff. Interviews were
conducted prior to the medical evaluation. An interviewer read the ques-
tionnaire items to children in a private room, and the children selected
answers from a response sheet. Approximately 2 weeks later, the family
was contacted by telephone in the evening on each of 5 consecutive school
days and an interviewer administered the Daily Diary Interview (DDI) to
the children (Walker, Smith, Garber, Van Slyke, & Claar, 2001). Three
months following the clinic visit, measures of somatic symptoms, depres-
sive symptoms, and functional disability were administered to the children
by telephone. The administration protocol was the same for all partici-
pants.4The clinic interview and diary interviews took 30–45 min, and the
follow-up interview (at 3 months) took 20 min. Children received $10 for
each assessment. The study was approved by the Institutional Review
Pain Beliefs Questionnaire (PBQ).
sessed with the PBQ (Van Slyke, 2001). The PBQ was developed to assess
appraisals (cf. Lazarus & Folkman, 1984; Smith & Lazarus, 1990) of pain
in children with recurrent pain. These beliefs are assumed to be relatively
stable. The PBQ includes conceptually derived subscales, each with six
items, to assess PFCP (e.g., “When I have a bad stomach ache, there are
ways I can get it to stop”) and EFCP (e.g., “I know I can handle it no matter
how bad my stomach hurts”). Children use a 5-point rating scale to indicate
how true each statement is about their abdominal pain. In this sample,
alpha reliabilities were .82 and .76 for PFCP and EFCP, respectively.
Children’s pain beliefs were as-
3Of the 21 participants with incomplete data, 11 did not report an
episode of pain during the week of diary interviews, and 10 were lost to
follow-up because of either inability to contact the family (e.g., due to
disconnected telephones) or because the family declined to participate
because of inadequate time available.
4Administration of measures at the clinic and by telephone was de-
signed to maximize the equivalence of the procedures. In both instances,
the interviewer read the questions to the child and the child selected
responses from a printed sheet that listed response options for each mea-
sure (these response sheets were mailed to children prior to the telephone
relations between short- and long-term outcomes were expected to be weak.
Specific model of hypothesized relations among variables. The dashed lines indicate that the depicted
TESTING A MODEL OF PAIN APPRAISAL AND COPING
Pain Response Inventory (PRI).
children’s typical styles of coping with abdominal pain. The PRI consists
of 60 items rated on a 5-point scale. It yields three broad-band factor
scores: Active Coping reflects problem-focused strategies aimed at pain
reduction (e.g., “Try to do something to make it go away”), Passive Coping
reflects strategies that avoid confronting pain (e.g., “Not even try to do
anything about it because it will not help”), and Accommodative Coping
reflects efforts to accept and adjust to pain (e.g., “Try to learn to live with
it”). Alpha reliabilities were .84 for Active Coping, .90 for Passive Coping,
and .88 for Accommodative Coping.
Children’s Somatization Inventory (CSI).
Zeman, 1991; Walker & Garber, 2003; Walker, Garber & Greene, 1991)
assesses the severity of nonspecific somatic symptoms (e.g., “headaches,”
“dizziness”) that often are reported by children with RAP and need not
have organic disease etiology (Walker et al., 1991). Respondents rate the
extent to which they have experienced each of 35 symptoms during the last
2 weeks using a 5-point scale, ranging from 0 (not at all) to 4 (a whole lot).
Three-month test–retest Pearson product–moment correlation reliabilities
for the CSI are .50 for well patients and .66 for patients with chronic pain
(Walker et al., 1991). In this sample, the CSI had an alpha reliability of .90
and a 3-month stability of .54.
Functional Disability Inventory (FDI).
1991) assessed children’s self-reported difficulty in physical and psycho-
social functioning due to their physical health during the past 2 weeks. The
FDI has high levels of internal consistency and 3-month test–retest reli-
ability estimates exceeding .60 for patients with RAP (Walker & Greene,
1991). Scores on the FDI correlate significantly with school absence
(Walker & Greene, 1991). In this sample, the FDI had an alpha reliability
of .88 and a 3-month stability of .56.
Children’s Depression Inventory (CDI).
assessed with the CDI (Kovacs, 1981; Kovacs & Beck, 1977). The CDI
contains 27 self-report items representing depressive symptoms, each rated
on a 3-point scale, and summed to obtain a total score. Reliability and
validity are adequate (Saylor, Finch, Spirito, & Bennett, 1984). In the
present sample, the CDI had an alpha reliability of .85 and a 3-month
stability of .74.
Episode-specific measures were obtained with the DDI
(Walker et al., 2001), administered to children in the evening on each of 5
school days. The DDI discriminates between pain patients and well chil-
dren and yields reliable measures of episode-specific pain appraisals, pain
coping behavior, and pain outcomes (Walker et al., 2001). The DDI asks
children to respond to structured questions regarding the worst pain epi-
sode of the day. Data from the first pain episode reported by the child
during the week were used to represent the episode-specific appraisals,
coping activities, and outcomes for that child. Measures related to these
pain episodes are described below.
Children’s appraisals of abdominal pain episodes were assessed with
respect to PFCP and EFCP. To assess perceived PFCP, we had children
first respond yes or no to the question, “When your stomach hurt, did you
think you would be able to do something to make it feel better?” Next,
children responded to the question, “How sure were you when your
stomach hurt that you (could–could not) do something to make it feel
better?” using a 5-point scale, ranging from 0 (not at all) to 4 (a whole lot).
The total score for PFCP was obtained by combining the responses to the
two questions into a single score that ranged from –4 (indicating high
degree of certainty that the child would not be able to reduce the pain) to
?4 (indicating a high degree of certainty that the child would indeed be
able to reduce the pain). A response of 0 on either of the two original scales
was coded as a 0 (i.e., at the midpoint) of this combined scale, resulting in
a single 9-point scale. Assessment of EFCP used a similar format with the
questions reworded to read, “When your stomach hurt, did you think you
would be able to deal with your pain or handle your pain, even if it did not
go away?” and “How sure were you that you (could–could not) deal with
it or handle it, even if your stomach pain did not go away?”
The PRI (Walker et al., 1997) assessed
The CSI (Garber, Walker,
The FDI (Walker & Greene,
Depressive symptoms were
Three types of episode-specific coping were assessed: Active Coping,
Passive Coping, and Accommodative Coping. Items were derived from the
PRI (Walker et al., 1997). Active Coping was assessed with three items:
“Try to figure out what to do about it” “Ask someone for help” and “Talk
to someone who you thought would understand how you felt.” Passive
Coping was assessed with three items: “Think to yourself that there was
nothing you could do, so you did not even try” “Go off by yourself ” and
“Think to yourself that the situation was going to get worse.” Accommo-
dative Coping was assessed with four items: “Try to accept it” “Think of
things to take your mind off the situation” “Tell yourself that the situation
was not that bad” and “Tell yourself to keep going even though this was
happening.” Children reported use of each strategy on a 5-point scale,
ranging from 0 (not at all) to 4 ( a whole lot). Mean alpha reliabilities for
the week were .70, .57, and .79 for Active Coping, Passive Coping, and
Accommodative Coping, respectively.
Episode-specific somatic symptoms were assessed with the Index of
Somatic Symptoms (Walker et al., 2001), an abbreviated state version of
the CSI (Garber et al., 1991; Walker et al., 1991) comprised of five
frequently endorsed items from the CSI (i.e., “headache,” “nausea,”
“tired,” “sore muscles,” “feeling weak”). Children reported how much they
felt each symptom during the pain episode. The mean alpha reliability for
the index was .73 across administrations.5
Episode-specific emotional distress was assessed with an abbreviated
state version of the Positive and Negative Affect Scale (Watson, Clark, &
Tellegen, 1988). Children reported how much they experienced each affect
during the pain episode on a 5-point scale, ranging from 0 (not at all) to 4
(a whole lot). Total scores for Positive Affect and Negative Affect were
constructed by summing responses to items corresponding to each sub-
scale. Episode-Specific Emotional Distress was constructed by subtracting
total Positive Affect from total Negative Affect. The mean alpha reliability
for the 12-item index was .81 across the week of administrations.
Overview of Model Testing and Preliminary Analyses
The proposed model of pain coping was tested using path
analysis based in structural equation modeling. This type of ap-
proach was selected because it simultaneously examines multiple
hypothesized paths of direct and indirect influence and can provide
global indices of the fit between the data and a proposed theoret-
ical model (Holmbeck, 1997; Peyrot, 1996). We examined the
interrelations among observed indicators and did not use multiple
indicators to model interrelations among latent constructs. This
limitation was imposed because the theoretical model we set out to
test (see Figure 2) is rather complex and the added complications
of modeling the latent constructs would have rendered the planned
model testing computationally unfeasible. All analyses were con-
ducted with the EQS program (Version 5.4; Bentler & Wu, 1995)
using the maximum-likelihood method of parameter estimation.
In presenting the path analyses, we report several indicators of
fit. The most basic indicator is a chi-square reflecting the degree of
5The original version of the Index of Somatic Symptoms (cf. Walker at
al., 2001) included a seventh item, “stomach ache.” This item was omitted
in the present study because our explicit focus was on an episode involving
a stomach ache, and thus ratings on this item could be presumed to be high.
Moreover, in this study, we were primarily interested in the degree to
which the abdominal pain episode was accompanied by additional somatic
WALKER, SMITH, GARBER, AND CLAAR
discrepancy between the observed covariance matrix derived from
the data and that predicted by the model. A small, nonsignificant
chi-square indicates that one cannot reject the null hypothesis that
the tested model fits the data. However, several problems are
associated with the chi-square statistic (see, e.g., Hu & Bentler,
1998). First, at a conceptual level, models are taken to be approx-
imations of reality, and testing whether the observed and predicted
covariance matrices are identical may be too strict a criterion.
Moreover, the chi-square statistic is dependent on sample size and
sensitive to model complexity and deviations from multivariate
normality in the data (e.g., Hu & Bentler, 1998; La Du & Tanaka,
1989). Accordingly, we report additional fit indices that circum-
vent these problems. First, the ratio of chi-square:degree of free-
dom takes model complexity into account. Values less than 3
reflect an acceptable model (e.g., Church & Burke, 1994; Marsh &
Hocevar, 1985). Second, the root-mean-square error of approxi-
mation (RMSEA) provides an estimate of the average absolute
discrepancy between the model covariance estimates and the ob-
served covariances. For this index, values less than .05 indicate a
close fit to the data, and values of about .08 represent an acceptable
fit (e.g., Browne & Cudek, 1993). Finally, the comparative fit
index (CFI; Bentler, 1990) indicates the degree to which the
theoretical model better fits the data than a base model constrain-
ing all constructs to be uncorrelated with one another. The CFI is
considerably more robust than the chi-square statistic from devi-
ations from multivariate normality. A CFI value above .90 reflects
a good fit to the data (e.g., Bentler, 1990).
Testing the Fit of the Hypothesized Model
Prior to testing the model, we examined bivariate correlations
among variables (see Table 1). The expected relations were gen-
erally observed. To examine the overall fit of the hypothesized
model, we initially tested the model as depicted in Figure 2. In
addition to the paths in the figure, the five exogenous variables
representing dispositional antecedents (beliefs and coping styles)
were allowed to intercorrelate with one another. The fit of this
model to the data was not acceptable, ?2(76, N ? 133) ? 353.4,
p ? .01, ?2:df ? 4.65, CFI ? .62, RMSEA ? .17. Accordingly,
several modifications were made based on a consideration of
theory and the modification indices yielded by the EQS algorithm.
Modifications to the a priori model were only made if they were
theoretically defendable and did not change the basic thrust of the
First, the error terms associated with the variables were allowed
to correlate within each of the three sets of variables representing
episode-specific appraisal, short-term outcomes, and long-term
outcomes. These changes allowed the model to account for the fact
that, perhaps because of shared method variance, the variables
within each of these three sets were more highly intercorrelated
than was allowed by the initial model.
Next, the constraints on how short-term outcomes could predict
long-term outcomes were loosened slightly. In the initial model,
episode-specific somatic symptoms were allowed to predict only
somatic symptoms and disability. In the revised model, episode-
Intercorrelations Among Variables Entering the Model
Variable123456789 1011 1213 1415
1. Dispositional problem-focused
2. Dispositional emotion-focused
3. Dispositional active coping
4. Dispositional accommodative
5. Dispositional passive coping
6. Episode-specific problem-
focused coping potential
7. Episode-specific emotion-
focused coping potential
8. Episode-specific active coping
10. Episode-specific passive
11. Episode-specific somatic
12. Episode-specific distress
13. Somatic symptoms as
14. Functional disability as
15. Depressive symptoms as
.29 .22 .11.12
?.03 .23.44 .19—
?.20 .01 .10 .43
?.20 .02 .12 .28.34 .26—
?.20 .04.05 .46
?.20.03 .04 .30.28.25 .77—
?.14.32 .33.37.53 .45—
Note. N ? 133. For all entries in the table, rs ? .17, ps ? .05; rs ? .22, ps ? .01.
TESTING A MODEL OF PAIN APPRAISAL AND COPING
specific somatic symptoms also were allowed to predict depressive
symptoms. Additionally, in the revised model, dispositional pas-
sive coping style was allowed to predict the three long-term
outcomes. This latter change reflects the relation between dispo-
sitional passive coping and long-term outcomes that has been
documented in previous pain literature (e.g., Brown et al., 1989;
Gil et al., 1991). In allowing these paths, we are indicating that, as
anticipated by the dashed lines in the original theoretical model
(see Figures 1 and 2), the trait-level relations between coping style
and long-term outcomes cannot fully be accounted for by the
observation of a single pain episode.
Finally, to simplify model presentation, we eliminated nonsig-
nificant paths among variables and nonsignificant correlations
among exogenous variables. The substantive paths of the resulting
model are depicted in Figure 3. Significant correlations among the
exogenous variables are depicted in Table 2, and those among the
relevant endogenous variables’ error terms are depicted in Table 3.
The fit of this modified model was adequate, ?2(75, N ? 133) ?
142.5, p ? .01, ?2:df ? 1.90, CFI ? .91, RMSEA ? .08. Overall,
this model conforms well to the original model depicted in Fig-
ure 2. For instance, each of the episode-specific appraisal and
coping variables was reliably predicted by its dispositional coun-
terpart, and in each case, the dispositional counterpart was the only
dispositional variable predicting the episode-specific variable.6In
addition, episode-specific data were consistent with the proposed
causal flow, such that appraisals of coping potential influenced
episode-specific coping activities, which in turn influenced short-
Nonetheless, there were several discrepancies between the pro-
posed and final models, especially concerning episode-specific
active coping. Contrary to expectations, active coping with the
pain episode was not predicted by appraisals of PFCP, was not
6The fact that this model fit was achieved without adding paths from
additional dispositional variables to the episode-specific appraisal and
coping variables suggests reasonable predictive and discriminant validity
among the appraisal and coping constructs. However, as a more direct test
of the discriminant validity within the appraisal and coping constructs,
respectively, two follow-up analyses were performed. In both analyses, the
final model was retested, but in the first analysis, episode-specific EFCP
was predicted by dispositional PFCP and episode-specific PFCP was
predicted by dispositional EFCP, and in the second analysis, each of the
three episode-specific coping strategies was predicted by the two disposi-
tional coping strategies other than its counterpart (e.g., episode-specific
accommodative coping was predicted by dispositional active and passive
coping but not by dispositional accommodative coping). In both analyses,
the modified models showed evidence of fitting less well than the final
theoretically based model: for the model with the switched predictors of
appraised coping potential, ?2(75, N ? 133) ? 167.1, p ? .01, ?2:df ?
2.23, CFI ? .87, RMSEA ? .10, and for the one with the altered predictors
of episode-specific coping, ?2(72, N ? 133) ? 202.2, p ? .01, ?2:df ?
2.81, CFI ? .82, RMSEA ? .12. These results provide further support for
the discriminant validity of the appraisal and coping constructs.
fit index ? .91, root-mean-square error of approximation ? .08. The numerical values represent standardized
beta weights. The dashed lines indicate that the depicted paths between short- and long-term outcomes were
expected to be weak.
Final model of observed relations among variables, ?2(75, N ? 133) ? 142.5, p ? .01, comparative
WALKER, SMITH, GARBER, AND CLAAR
related to episode-specific emotional distress, and was positively,
rather than negatively, associated with episode-specific somatic
symptoms. These results are in line with a history of inconsistent
results for active or problem-focused modes of coping (cf. Smith,
Wallston, & Dwyer, 2003). Other findings that deviate from the
proposed model concern episode-specific accommodative coping,
which was predicted by appraisals of EFCP, as hypothesized, but
was not associated with lower appraisals of PFCP. In turn,
episode-specific accommodative coping was associated with lower
levels of short-term emotional distress, as predicted, but did not
predict lower levels of somatic symptoms.
Testing a Follow-Up Model
We hypothesized that short-term outcomes of the pain episode
would be related to long-term outcomes but these relations would
be modest because of the limited impact that a single pain episode
should have on long-term adjustment. In fact, statistically reliable,
if modest (standardized ?s ? .16?.25), relations to the long-term
outcomes were found: Episode-specific somatic symptoms were
associated with all three outcomes, and episode-specific emotional
distress was associated with depressive symptoms. However, the
meaning of these relations is ambiguous because the long-term
outcomes are relatively stable, as indicated by their 3-month sta-
bilities presented above. Thus, without controlling for prior levels
of the long-term outcomes, one cannot determine whether the
observed relations primarily reflect relations of the short-term
outcomes to the stable component of the long-term outcomes, or
whether they are associated with changes in the outcomes over
3-months, which would be suggestive of the hypothesized causal
contribution of the short-term outcomes to the long-term
To investigate this issue, we tested a final model in which
baseline values of the three long-term outcomes at the time of the
initial clinic visit were added to the model. These three additional
exogenous variables were allowed to correlate freely with one
another and with the other exogenous dispositional antecedents,
and each was allowed to predict to itself (only) at the follow-up
assessment. The overall fit of this model (not depicted) was also
adequate, ?2(98, N ? 133) ? 178.7, p ? .01, ?2:df ? 1.82, CFI ?
.93, RMSEA ? .08. Two aspects of this model are noteworthy.
First, other than the relations to the long-term outcomes, none of
the associations among variables depicted in Figure 3 were
changed appreciably by adding initial values of the long-term
outcomes to the model. This suggests that the relations observed in
the model are computationally stable. Second, the relations be-
tween the short-term and long-term outcomes were weakened and
no longer reached statistical significance. However, although the
relation between episode-specific emotional distress and long-term
depressive symptoms essentially disappeared (? ? .05, z ? 1), the
relations of short-term somatic symptoms maintained trend-level
associations to each of the long-term outcomes (?s ? .14, .12, and
.16; zs ? 1.84, 1.65, and 1.68 for somatic symptoms, functional
disability, and depressive symptoms, respectively). These latter
relations suggest that, in repeated occasions over time, the short-
term outcomes may influence long-term outcomes.
The theoretical literature on the development of chronic pain in
children has highlighted the importance of exaggerated threat
appraisal and passive coping in sustaining chronic pain (Walker,
1999; Zeltzer & Feldman, 1999). We have extended that literature
by differentiating threat appraisals into appraisals of PFCP and
EFCP, by examining accommodative coping in addition to active
and passive coping and by testing a model that includes both
dispositional and episode-specific pain appraisals and coping as
predictors of health outcomes. Our findings provide empirical
support and a more fine-grained picture of the process by which
children’s pain appraisals and coping may influence a range of
outcomes including somatic symptoms, depressive symptoms, and
Regarding passive coping, we found that children who believed
that they had little ability either to reduce or to accept pain were
likely to appraise their ability to cope with a specific pain episode
as low and to exhibit passive coping behavior in confronting that
pain episode. Passive coping with a pain episode, defined as social
withdrawal and catastrophizing cognitions, was associated with
higher levels of episode-specific symptoms and emotional distress.
These outcomes, in turn, had small but statistically reliable effects
on somatic symptoms, functional disability, and depressive symp-
toms assessed 3 months later. The latter effects, which were
predicted to be weak, were still evident as statistical trends, even
after controlling for initial values of the long-term outcomes. This
suggests that over repeated pain episodes, the influence of episode-
Intercorrelations Among Exogenous Variables
1. Dispositional problem-focused
2. Dispositional emotion-focused
3. Dispositional active coping style
4. Dispositional accommodative
5. Dispositional passive coping style
Empty cells represent nonsignificant correlations.
All depicted correlations are statistically significant at p ? .05.
Intercorrelations Among Selected Model Error Terms
Variable EEFCE EDISE OSYME OFDE
Dashes indicate that correlations were not computed. EPFCE ? error term
for episode-specific problem-focused coping potential; ESYME ? error
term for episode-specific somatic symptoms; OFDE ? error term for
functional disability as an outcome; ODEPE ? error term for depressive
symptoms as an outcome; EEFCE ? error term for episode-specific
emotion-focused coping potential; EDISE ? error term for episode-
specific distress; OSYME ? error term for somatic symptoms as an
All depicted correlations are statistically significant at p ? .05.
TESTING A MODEL OF PAIN APPRAISAL AND COPING
specific outcomes on long-term outcomes could become profound.
Finally, consistent with the notion that passive coping fuels an
escalating cycle of pain amplification (Philips, 1987), we found
that dispositional passive coping had significant direct effects both
on episode-specific passive coping and on long-term symptoms
Accommodative coping with pain, defined as acceptance of pain
and self-encouragement, has been less extensively studied than
either passive or active coping with pain. In this study, appraisals
of higher EFCP predicted episode-specific accommodative coping,
which was associated with lower levels of emotional distress
related to the pain episode. Lower episode-specific distress, in
turn, predicted lower levels of depressive symptoms 3 months
later. Thus, accommodative coping, which can be conceptualized
as a potentially adaptive type of emotion-focused (Lazarus &
Folkman, 1984; Smith & Lazarus, 1990) or secondary control
engagement coping (Thomsen et al., 2002), appears to be relevant
for the regulation of both short- and long-term negative affect. To
the extent that accommodative coping facilitates children’s expo-
sure to situations associated with pain, it may help break the
escalating cycle of pain fear and avoidance that characterizes
chronic pain (Bursch et al., 1998).
The unexpected findings for episode-specific active coping may
be explained by several factors. First, the fact that the relation of
dispositional active coping to episode-specific active coping was
lower than the relation of passive or accommodative coping to
their episode-specific counterparts suggests that active coping may
be more dependent on environmental resources and thus more
context specific than other types of coping. Moreover, the meaning
of active coping may depend on the nature of concurrent coping
strategies. For example, the impact of seeking social support (an
active coping strategy) is likely to differ when combined with
self-encouragement than with catastrophizing cognitions. In the
latter case, seeking social support might actually represent a pas-
sive strategy for attempting to get someone else to take responsi-
bility for a problem. Finally, it is possible that active coping
represents targeted problem-solving efforts by some children but
in other cases reflects indiscriminant efforts to try anything. For
example, children experiencing high levels of distress may engage
in a variety of active coping efforts that constitute a cry for help.
Thus, the small but significant positive relation between episode-
specific active coping and episode-specific somatic symptoms may
reflect a direction of causality opposite that hypothesized in our
This study is limited by reliance on self-report measures and by
the fact that the episode-specific data were embedded in a longi-
tudinal design but were themselves cross-sectional. Although the
data proved to be largely consistent with the model we hypothe-
sized and tested, it is possible to generate alternative models that
fit the data just as well. As is generally the case with cross-
sectional data, changing the direction of influence for supported
relations typically will not change a model’s fit. Thus, it is possible
that episode-specific outcomes could influence episode-specific
coping or that the relations could be bidirectional. For instance,
increases in symptoms could lead to increased use of passive
coping strategies instead of, or in addition to, the reverse. To sort
out such issues, future researchers should use observations of
children’s coping with pain in laboratory and natural environ-
ments. Replication of this study in other populations of pediatric
pain patients, including those with organic disease, would allow a
test of the generalizability of the model. In studies of adult pain
patients, Turk and Rudy (1988) found that profiles of pain coping
were similar across pain populations. This also may be the case for
pediatric pain and would have implications for the development of
pain management interventions with broad applicability.
Another limitation of this study is that the model was tested on
a fairly small sample and has not been cross-validated. A primary
concern with small samples is that the observed correlation coef-
ficients have relatively large standard errors. The imprecision in
these estimated correlations can cause conditions (such as the
correlation matrix failing to be positive definite) that prevent the
modeling algorithms from converging on a solution or that lead the
algorithms to converge on unstable, nonsensical solutions. The
likelihood of these outcomes increases as model complexity in-
creases. A secondary concern is that the parameter estimates and
fit indices are derived on the basis of asymptotic models, and thus
they are less accurate with smaller samples (see Bollen, 1989).
We took steps to avoid these problems: We limited the com-
plexity of our initial model by making several simplifying assump-
tions, we modeled observed indicators rather than latent con-
structs, and we relied primarily on fit indices, such as the CFI and
RMSEA, that are less sensitive to sample size and model com-
plexity than is the chi-square. Moreover, the model we evaluated
appeared to be both computationally stable (in that individual
parameters did not change greatly when various paths and vari-
ables were added or removed from the model) and theoretically
meaningful. Thus, we are confident that the final model is sound,
despite the relatively small sample on which it was based. None-
theless, as would be the case even with a larger sample, it is
important to cross-validate this model in subsequent work.
It also should be noted that in administering the PBQ and the
PRI only at the initial clinic visit, we assumed that our disposi-
tional constructs were relatively stable over time. The predictive
relations that the dispositional appraisal and coping variables dem-
onstrated to their episode-specific counterparts assessed 2 weeks
after the clinic visit offer some validation to this assumption.
Nonetheless, the stability of these measures needs to be explicitly
Finally, the model tested in this study does not consider con-
textual factors that may impact children’s pain behavior (cf.
McGrath, 1994). The modest, albeit significant, relations we ob-
served between dispositional appraisals and coping and their
episode-specific counterparts suggest that contextual factors may
combine with children’s dispositional response tendencies to de-
termine coping in response to specific episodes of pain. Research
on how contextual factors, such as parent responses (cf. Frank,
Blount, Smith, Manimala, & Martin, 1995; Walker, Claar, &
Garber, 2002), and dispositional factors interact to produce coping
will provide useful information for the design of interventions to
enhance children’s pain coping skills. Our findings further suggest
that accommodative strategies, such as acceptance and positive
reappraisal, may be useful in regulating children’s negative affect
associated with chronic pain. Thus, in future research on children’s
coping with chronic pain, it will be important to differentiate
coping strategies beyond the dichotomy of active and passive
strategies and to consider the potentially beneficial role of accom-
modative coping strategies.
WALKER, SMITH, GARBER, AND CLAAR
Apley, J. (1975). The child with abdominal pains. London: Blackwell.
Bentler, P. M. (1990). Comparative fit indexes in structural models. Psy-
chological Bulletin, 107, 238–246.
Bentler, P. M., & Wu, E. J. C. (1995). EQS for Macintosh: User’s guide
[Computer software manual]. Encino, CA: Multivariate Software.
Bollen, K. A. (1989). Structural equations with latent variables. Oxford,
Brown, G. K., & Nicassio, P. M. (1987). Development of a questionnaire
for the assessment of active and passive coping strategies in chronic pain
patients. Pain, 31, 53–63.
Brown, G. K., Nicassio, P. M., & Wallston, K. A. (1989). Pain coping
strategies and depression in rheumatoid arthritis. Journal of Consulting
and Clinical Psychology, 57, 652–657.
Browne, M. W., & Cudek, R. (1993). Alternative ways of assessing model
fit. In K. A. Bollen & J. S. Long (Eds.), Testing structural equation
models (pp. 136–162). Newbury Park, CA: Sage.
Bursch, B., Walco, G. A., & Zeltzer, L. (1998). Clinical assessment and
management of chronic pain and pain-associated disability syndrome.
Developmental and Behavioral Pediatrics, 19, 45–53.
Carver, C. S., & Scheier, M. F. (1994). Situational coping and coping
dispositions in a stressful transaction. Journal of Personality and Social
Psychology, 66, 184–195.
Church, A. T., & Burke, P. J. (1994). Exploratory and confirmatory tests of
the Big Five and Tellegen’s three- and four-dimensional models. Journal
of Personality and Social Psychology, 66, 93–114.
Claar, R. L., Walker, L. S., & Smith, C. A. (2002). The influence of
appraisals in understanding children’s experiences with medical proce-
dures. Journal of Pediatric Psychology, 27, 553–563.
Folkman, S. (1984). Personal control and stress and coping processes: A
theoretical analysis. Journal of Personality and Social Psychology, 46,
Folkman, S., Lazarus, R. S., Dunkel-Schetter, C., DeLongis, A., & Gruen,
R. J. (1986). Dynamics of a stressful encounter: Cognitive appraisal,
coping, and encounter outcomes. Journal of Personality and Social
Psychology, 50, 992–1003.
Frank, N. C., Blount, R. L., Smith, A. J., Manimala, M. R., & Martin, J. K.
(1995). Parent and staff behavior, previous child medical experience,
and maternal anxiety as they relate to child procedural distress and
coping. Journal of Pediatric Psychology, 20, 277–289.
Garber, J., Walker, L. S., & Zeman, J. L. (1991). Somatization symptoms
in a community sample of children and adolescents: Further validation
of the Children’s Somatization Inventory. Psychological Assessment, 3,
Gil, K. M., Williams, D. A., Thompson, R. J., & Kinney, T. R. (1991).
Sickle cell disease in children and adolescents: The relation of child and
parent pain coping strategies to adjustment. Journal of Pediatric Psy-
chology, 16, 643–663.
Holmbeck, G. N. (1997). Toward terminological, conceptual, and statisti-
cal clarity in the study of mediators and moderators: Examples from the
child–clinical and pediatric psychology literatures. Journal of Consult-
ing and Clinical Psychology, 65, 599–610.
Hu, L., & Bentler, P. M. (1998). Fit indices in covariance structure
modeling: Sensitivity to underparameterized model misspecification.
Psychological Methods, 4, 424–453.
Jensen, M. P., Turner, J. A., & Romano, J. M. (1991). Self-efficacy and
outcome expectancies: Relationship to chronic pain coping strategies
and adjustment. Pain, 44, 263–269.
Kovacs, M. (1981). Rating scales to assess depression in school-aged
children. Acta Paedopsychiatrica, 46(Suppl. 5–6), 305–315.
Kovacs, M., & Beck, A. T. (1977). An empirical–clinical approach toward
a definition of childhood depression. In J. Schulterbrandt & A. Raskin
(Eds.), Depression in childhood: Diagnosis, treatment, and conceptual
models (pp. 1–25). New York: Raven Press.
La Du, T. J., & Tanaka, J. S. (1989). Influence of samples size, estimation
methods, and model specification on goodness-of-fit assessments in
structural equation modeling. Journal of Applied Psychology, 74, 625–
Lazarus, R. S. (1991). Emotion and adaptation. New York: Oxford Uni-
Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal, and coping. New
Lefebvre, J. C., Keefe, F. J., Affleck, G., Raezer, L. B., Starr, K., Caldwell,
D. S., & Tennen, H. (1999). The relationship of arthritis self-efficacy to
daily pain, daily mood, and daily pain coping in rheumatoid arthritis
patients. Pain, 80, 425–435.
Marsh, H. W., & Hocevar, D. (1985). Application of confirmatory factor
analysis to the study of self-concept: First- and higher order factor
models and their invariance across groups. Psychological Bulletin, 97,
McCracken, L. M. (1998). Learning to live with the pain: Acceptance of
pain predicts adjustment in persons with chronic pain. Pain, 74, 21–27.
McGrath, P. A. (1994). Pain in children: Nature, assessment, and treat-
ment. New York: Guilford Press.
Monroe, S. M., & Kelley, J. M. (1995). Measurement of stress appraisal.
In S. Cohen, R. Kessler & L. Gordon (Eds.), Measuring stress: A guide
for health and social scientists (pp. 122–147). New York: Oxford
Peyrot, M. (1996). Causal analysis: Theory and application. Journal of
Pediatric Psychology, 21, 3–24.
Philips, H. C. (1987). Avoidance behaviour and its role in sustaining
chronic pain. Behaviour Research and Therapy, 25, 273–279.
Ptacek, J. T., & Gross, S. (1997). Coping as an individual difference
variable. In G. R. Pierce et al. (Eds.), Sourcebook of social support and
personality (pp. 69–92). New York: Plenum/Kluwer Academic.
Rudolph, K. D., Dennig, M. D., & Weisz, J. R. (1995). Determinants and
consequences of children’s coping in the medical setting: Conceptual-
ization, review, and critique. Psychological Bulletin, 118, 328–357.
Saylor, C. F., Finch, A. J., Jr., Spirito, A., & Bennett, B. (1984). The
Children’s Depression Inventory: A systematic evaluation of psycho-
metric properties. Journal of Consulting and Clinical Psychology, 52,
Schwartz, J. E., Neale, J., Marco, C., Shiffman, S. S., & Stone, A. (1999).
Does trait coping exist? A momentary assessment approach to the
evaluation of traits. Journal of Personality and Social Psychology, 77,
Smith, C. A., & Lazarus, R. S. (1990). Emotion and adaptation. In L. A.
Pervin (Ed.), Handbook of personality: Theory and research (pp. 609–
637). New York: Guilford Press.
Smith, C. A., Wallston, K. A., & Dwyer, K. A. (2003). Coping and
adjustment to rheumatoid arthritis. In J. Suls & K. A. Wallston (Eds.),
Social psychological foundations of health and illness (pp. 458–494).
Malden, MA: Blackwell.
Smith, C. A., Wallston, K. A., Dwyer, K. A., & Dowdy, S. W. (1997).
Beyond good and bad coping: A multidimensional examination of
coping with pain in persons with rheumatoid arthritis. Annals of Behav-
ioral Medicine, 19, 11–21.
Thompson, R. J., Gil, K. M., Abrams, M. R., & Phillips, G. (1992). Stress,
coping, and psychological adjustment of adults with sickle cell disease.
Journal of Consulting and Clinical Psychology, 60, 433–440.
Thomsen, A. H., Compas, B. E., Colletti, R. B., Stanger, C., Boyer, M. C.,
& Konik, B. S. (2002). Coping and stress responses in children with
recurrent abdominal pain. Journal of Pediatric Psychology, 27, 215–
Turk, D. C., & Rudy, T. E. (1988). Toward an empirically derived taxon-
omy of chronic pain patients: Integration of psychological assessment
data. Journal of Consulting and Clinical Psychology, 56, 233–238.
Turner, J. A. (1991). Coping and chronic pain. In M. R. Bond, J. E.
TESTING A MODEL OF PAIN APPRAISAL AND COPING
Charlton, & C. J. Woolf (Eds.), Pain research and clinical management: Download full-text
Vol. 4. Proceedings of the 6th World Congress on Pain (pp. 219–227).
New York: Elsevier.
Van Slyke, D. A. (2001). Parent influences on children’s pain behavior.
Unpublished doctoral dissertation, Vanderbilt University.
Walker, L. S. (1999). The evolution of research on recurrent abdominal
pain: History, assumptions and new directions. In P. J. McGrath & G. A.
Finley (Eds.), Chronic and recurrent pain in children and adolescents
(pp. 141–172). Seattle, WA: International Association for the Study of
Walker, L. S., Claar, R. L., & Garber, J. (2002). Social consequences of
children’s pain: When do they encourage symptom maintenance? Jour-
nal of Pediatric Psychology, 27, 689–698.
Walker, L. S., & Garber, J. (2003). Manual for the Children’s Somatization
Inventory. Nashville, TN: Vanderbilt University Medical Center, De-
partment of Pediatrics.
Walker, L. S., Garber, J., & Greene, J. W. (1991). Somatization symptoms
in pediatric abdominal pain patients: Relation to chronicity of abdominal
pain and parent somatization. Journal of Abnormal Child Psychology,
Walker, L. S., Garber, J., & Greene, J. W. (1993). Psychosocial correlates
of recurrent childhood pain: A comparison of pediatric patients with
recurrent abdominal pain, organic illness, and psychiatric disorders.
Journal of Abnormal Psychology, 102, 248–258.
Walker, L. S., & Greene, J. W. (1991). The Functional Disability Inven-
tory: Measuring a neglected dimension of child health status. Journal of
Pediatric Psychology, 16, 39–58.
Walker, L. S., Smith, C. A., Garber, J., & Van Slyke, D. A. (1997).
Development and validation of the Pain Response Inventory for chil-
dren. Psychological Assessment, 9, 392–405.
Walker, L. S., Smith, C. A., Garber, J., Van Slyke, D. A., & Claar, R.
(2001). The relation of daily stressors to somatic and emotional symp-
toms in children with recurrent abdominal pain. Journal of Consulting
and Clinical Psychology, 69, 85–91.
Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and vali-
dation of brief measures of positive and negative affect: The PANAS
scales. Journal of Personality and Social Psychology, 54, 1063–1070.
Zeltzer, L., & Feldman, S. (1999). Soothing and chronic pain. In M. Lewis
& D. Ramsay (Eds.), Soothing and stress (pp. 195–227). Mahwah, NJ:
WALKER, SMITH, GARBER, AND CLAAR