Journal of Pediatric Psychology, Vol. 28, No. 8, 2003, pp. 579-588
© 2003 Society of Pediatric Psychology
Parent and Child Reporting of Negative Life Events: Discrepancy and Agreement across Pediatric Samples
1 University of Kansas, Memphis, Tennessee, 2 St. Jude Children's Research Hospital, Memphis, Tennessee
All correspondence concerning this article should be addressed to Ric Steele, PhD, Clinical Child Psychology Program, 2006 Dole Center for Human Development, University of Kansas, Lawrence, Kansas 66045–7555. E-mail: rsteele{at}ku.edu
 |
Abstract |
|---|
 Top Abstract IntroductionDiscrepancy in Parent and...Differences in Reporting of...Goals of the StudyMethodsResultsDiscussionReferences |
|---|
Objective To examine the consistency in child and parent reporting of child's negative life events across child/pediatric samples. Methods A total of 613 child-parent dyads provided independent reports of negative life events. The pairs included three groups consisting of children who were healthy (n = 362), diagnosed with cancer (n = 130), and diagnosed with a chronic illness (juvenile rheumatoid arthritis, diabetes, or cystic fibrosis; n = 121). Results Children reported significantly more negative life events than their parents reported for them. Additionally, children in the chronically ill group self-reported significantly fewer negative life events than the other groups. However, parents of children with cancer reported significantly more negative life events than the other groups. Although discrepancies exist in all three samples, parents and children in the healthy group were significantly more discrepant than the other groups. Conclusions These results suggest that communication of children's life events between parent and child may increase during children's experience of cancer or a chronic illness. However, significant discrepancies remain in child and parent report of negative life events. Because of this, clinicians are encouraged to recognize the strengths and limitations of using multiple reporters in assessing negative life events in children.
Key words: negative life events; parent-child agreement; child self-report; chronic illness.
|
Introduction |
|---|
|
TopAbstractIntroductionDiscrepancy in Parent and...Differences in Reporting of...Goals of the StudyMethodsResultsDiscussionReferences |
|---|
Negative life events have been associated with many consequences in children, including an increase in behavior problems (MacLean, Perrin, Gortmaker, & Pierre, 1992
), a decrease in physical and psychological functioning
(Gersten, Langner, Eisenberg, &
Simcha-Fagan, 1977; Heisel,
Ream, Raitz, Rappaport, & Coddington, 1973;
Sandler & Ramsey, 1980),
and an overall decrease in adjustment in both school and home settings
(Hodges, London, & Colwell,
1990). Examples of such negative life events include divorce,
severe accidents or illnesses, and problems with school.
Bailey and Garralda (1990)
reported that the majority of studies examining the impact of stressful life
events of children have used primarily parent report as the method for data
collection. Despite the preponderance of studies utilizing parent report,
child self-report of negative life events has also been used with acceptable
reliability and validity (Jackson &
Warren, 2000; Johnson &
McCutcheon, 1980; Kager &
Holden, 1992). When child or adolescent reports are examined,
negative life events have been shown to predict internalizing disorders
(Vinnick & Erickson,
1992), substance abuse (Dinges
& Duong-Tran, 1992), and number of school days missed
(Swearingen & Cohen,
1985).
|
Discrepancy in Parent and Child Report |
|---|
|
TopAbstractIntroductionDiscrepancy in Parent and...Differences in Reporting of...Goals of the StudyMethodsResultsDiscussionReferences |
|---|
One finding that has emerged from the literature involves the discrepancy between parent and child reports of negative life events. This discrepancy has greatest relevance in clinical settings, where the use of both child and parent report may impact areas to be addressed in therapy. In fact, Yeh and Weisz (2001
) reported that
parents and children demonstrated poor agreement regarding the child's target
problem being addressed in therapy and that agreement between parents and
children was lowest for life stressors. Although current data do not indicate
the reason for the poor agreement of the parent-child pair, Yeh and Weisz
(2001) hypothesized that
parents and children do not effectively communicate about the reason the child
is in treatment. They further encouraged future research to address possible
influences on parent-child discrepancy.
Specific discrepancies for children's negative life events have been shown
to exist between child and parent report, including differences in the number
of events reported, the impact of events, and the reported onset of events.
For example, Bailey and Garralda
(1990) demonstrated that
children report significantly more negative life events than their parents
report for them. Similarly, Rende and Plomin
(1991) found that children
reported a greater frequency of events than their parents. However, parental
ratings of life event–related stress for their child were significantly
higher than their child's ratings of stress. Sandberg, Rutter, Pickles,
McGuinness, and Angold (2001)
also demonstrated that parent-reported onset of life events was statistically
different than child-reported onset. One hypothesized explanation of these
findings is that children's episodic memory is relatively immature,
compromising their abstractions of timing
(Bailey & Garralda, 1990).
Because of potentially compromised appreciation of time, children's reports of
negative life events may differ from their parents', particularly when the
occurrence of events to a specific time frame (e.g., the last 6 months) is
queried. To date, investigations of discrepancies between parent and child
reports of negative life events have not included multiple time frames as an
objective test of this hypothesis.
|
Differences in Reporting of Life Events by Pediatric Samples |
|---|
|
TopAbstractIntroductionDiscrepancy in Parent and...Differences in Reporting of...Goals of the StudyMethodsResultsDiscussionReferences |
|---|
In addition to the discrepancy found between parent and child reporting of negative life events, frequency of life events has also been shown to differ as a function of different child illnesses or medical conditions. For example, using parent report, Jacobs and Charles (1980
) found that certain
negative life events (e.g., marital separation, death of a family member)
occurred more frequently for children with cancer than healthy children. Yet
Bailey and Garralda (1990)
reported that children with chronic illnesses other than cancer reported
fewer negative life events than healthy children. This may suggest
that childhood cancer is unique in comparison with other chronic illnesses, or
it may reflect different outcomes as a function of informant: parents in the
Jacobs and Charles (1980)
study, and children in the study by Bailey and Garralda
(1990). Although these findings
might suggest that children with cancer would report more negative life events
than children with other chronic illnesses, such comparisons have not been
conducted.
Researchers have suggested a causal link between the reported occurrence of
negative childhood life events among these samples and subsequent health
problems (e.g., cancer; Forsen,
1991; Mandal, Ghosh, &
Nair, 1992), and similar explanations have been offered for the
higher self-reported negative life events among pediatric cancer patients
(e.g., Jacobs & Charles,
1980). While this is certainly a possibility, other explanations
also have merit, e.g., individuals with recently diagnosed cancer may be more
likely to recall negative events than individuals with no recent diagnosis.
This explanation is also consistent with Bailey and Garralda's
(1990) findings, in that
children with lifelong chronic illnesses (e.g., cystic fibrosis, as opposed to
children with cancer) would have no recent diagnosis to impact their
recall. To date, no examination of self-reported negative life events has been
conducted across illness groups.
|
Goals of the Study |
|---|
|
TopAbstractIntroductionDiscrepancy in Parent and...Differences in Reporting of...Goals of the StudyMethodsResultsDiscussionReferences |
|---|
The present investigation was designed around three primary goals. First, we examined the discrepancy in the reporting of negative life events by parent-child dyads, in order to replicate previous findings. Consistent with Bailey and Garralda (1990
), we
hypothesized that children would self-report more negative life events than
their parents report for them.
Second, we compared the number of negative life events across groups of
children with various health/illness conditions. Specifically, we included
healthy children, children on active therapy for one of three different
chronic illnesses, and children on active therapy for cancer. Consistent with
Mandal et al. (1992) and
Jacobs and Charles (1980), we
hypothesized that children in the cancer group and their parents would report
significantly more negative life events than the healthy group. Furthermore,
based on results by Bailey and Garralda
(1990), we hypothesized that
children with other chronic illnesses would report significantly fewer
negative life events than healthy children or children with cancer. While some
investigations have examined children with medical conditions relative to
healthy children, this is the first examination of child- and parent-reported
negative life events among groups of children with different illnesses.
Finally, we examined the reporting of life events across two specific time
frames (i.e., events over the past year and events over the lifetime of the
child). Previous investigations have examined negative life events in only a
single time period (e.g., 3 months; Bailey
& Garralda, 1990). Because of the likelihood of inaccuracies
in children's memories for the timing of events
(McCormack & Hoerl, 1999),
we hypothesized that parent-child agreement will be higher for events
occurring over the child's lifetime relative to events in the past year. That
is, we expect higher agreement that an event occurred, and lower
agreement that the event occurred within the past year.
|
Methods |
|---|
|
TopAbstractIntroductionDiscrepancy in Parent and...Differences in Reporting of...Goals of the StudyMethodsResultsDiscussionReferences |
|---|
Participants
Negative life events were examined among 613 parent and child dyads. However, 85 dyads were excluded because the relationship of the caretaker to the child was unknown. Only caretakers reported as mothers or fathers were retained for this study. An additional 12 participants were excluded because of incomplete data. For the analyses, a total of 66 fathers and 430 mothers were included in this sample.
Children were divided into three groups: those on active treatment for long-term chronic illnesses (n = 91), those on active treatment for cancer (n = 116), and healthy children with no known chronic or serious illness (n = 287). The chronically ill group comprised children diagnosed with cystic fibrosis (CF), diabetes mellitus (DM), and juvenile rheumatoid arthritis (JRA). Children with these diagnoses were grouped together for two reasons. First, CF, DM, and JRA are long-term illnesses that are usually diagnosed early in a child's life, whereas the children with cancer received their diagnoses more recently. Second, preliminary analyses indicated no differences in the total number of events endorsed between the subsamples in the chronically ill group, F(2, 82) = 1.84, ns. Children ranged in age from 7 to 18 years.
Participants in the chronically ill group were contacted at their respective specialty clinics at a local pediatric hospital during a routine office visit. Participants in the cancer group were contacted at routine clinic visits during the active stage of therapy at a large children's hospital with a national catchment area. Parental consent was obtained by graduate research assistants contacting families at random from a list of eligible patients. A total of 126 children with cancer and their parents were approached, and 116 (92%) agreed to participate. A total of 100 children with a chronic illness and parents were approached, and 91 (91%) agreed to participate.
The healthy control sample was recruited from four schools (two public and two private) from the same metropolitan area as the chronically ill and cancer samples. Letters explaining the purpose and requirements of the study were distributed by teachers among designated second- to twelfth-grade classes and sent home for parental consent. Students returning the letter with parental consent were considered eligible for the study. In addition to providing informed consent, parents were asked to confirm that their child was in good health and not undergoing treatment for any significant illness. Data were obtained from 287 healthy children, representing just under half (43%) of the 670 letters distributed. This represents an approximate rate of participation because an exact rate of refusals was not available due to student absences. The purpose and requirements of the investigation were explained to all parents (i.e., parents of children in the cancer, chronically ill, and healthy groups), and informed consent was obtained in accordance with institutional review board and American Psychological Association guidelines.
For the total sample, the mean age was 12.64 years (SD = 2.96)
with 206 male (42%) and 288 female children. Of the child participants, 403
(81%) were white, 72 (15%) were African American, and 17 (3%) were from other
racial and ethnic backgrounds. The socioeconomic status (SES) of the families,
as measured by the Hollingshead
(1975) four-factor index,
ranged from major business/professional to unskilled laborers, with a majority
of the families falling in the middle-class range. For the present analyses,
socioeconomic categories were collapsed into three levels (i.e., lower income,
middle income, and higher income). Demographic variables for each group of
children are listed in Table
I.
|
Instruments
A modified version of the Coddington Life Events Questionnaire (CLEQ;
Coddington, 1972) was used to
assess life events experienced by the child. This measure was chosen because
of the relevance of the specific item content. However, two significant
modifications were made. First, because we were specifically interested in the
reporting of negative life events, we excluded items that referred to
positive events (e.g., "Outstanding personal achievement").
Second, some items on the Coddington focus specifically on events relating to
medical illness (e.g., "Serious illness requiring
hospitalization"). Because inclusion of these items would necessarily
differ across groups in our sample, items pertaining to the target child's
illness were removed. Our modified version of the instrument consisted of 22
items that focused on major negative events (see
Table II). All items required
only a "yes" or "no" response. Each item required two
responses: once for whether the event had occurred at any point in the child's
life, and once for whether the event had occurred in the past year.
Scale-level reliability statistics (e.g., coefficient alpha, spit-half
reliability) usually are not reported for the Coddington, since individual
items would not be expected to covary with one another.
|
Procedure
For healthy children, measures were administered in a group format in their
classrooms during regular school hours. In elementary school classes, the
items were read aloud by research staff. In middle and high school classes,
the measures were distributed, and students completed them independently
during designated class time set aside for this task. Directions were
presented orally, and participants were allowed to work at their own pace. A
research assistant was available to answer questions. Parents of children in
the healthy group completed a modified version of the CLEQ with instructions
to complete the information regarding their child's experiences and return the
measure via mail.
For children in the chronically ill group, research assistants individually administered the self-report measures during routine clinic visits, and for participants in the cancer group, research assistants individually administered the self-report measures to the participants during clinic visits while the child was on active therapy. Parents and children in the chronically ill and cancer groups were administered the measures independently. All parents and children were administered the same version of the CLEQ with instructions to complete the information regarding the child's experiences. These procedures, and all aspects of participant recruitment, data collection, and data management were approved by the institutional review board.
|
Results |
|---|
|
TopAbstractIntroductionDiscrepancy in Parent and...Differences in Reporting of...Goals of the StudyMethodsResultsDiscussionReferences |
|---|
Analyses were organized by four study questions. First, demographic variables (i.e., age, race, SES, and gender) were examined to determine their influence on the reporting of negative life event. Second, analyses were conducted to determine whether children reported a different number of negative life events than their parents. Next, we compared the number of events reported by children and their parents across the three groups (i.e., cancer, chronically ill, healthy). Finally, the percent agreement of child and parent reports were compared across the three groups.
Demographic Variables
A total of three multivariate analyses of variance (MANOVAs) were conducted
to determine the impact of the demographic variables of race, age, and SES on
the reporting of negative life events. First, three groups (i.e., white,
African American, and other) were compared based on race. The omnibus MANOVA
was significant, F(4, 946) = 3.82, p < .01. Univariate
analysis of variance (ANOVA) indicated that children's reporting of life
events was significant across race for both child, F(2, 474) = 7.27,
p < .001, and adult report, F(2, 474) = 4.75, p
< .01. Specifically, post hoc analyses (Tukey's least significant
difference [LSD]) indicated that white children reported significantly fewer
negative life events than African American children (p < .05).
Similarly, parents of African American children reported significantly more
negative life events than parents of white children (p < .05).
Next, three groups (i.e., children, preadolescents, and adolescents) were compared based on the child's age. Age was also significantly related to the reporting of negative life events, F(4, 944) = 5.12, p < .001. Univariate ANOVA indicated that children's reporting of life events was significantly different across three age groups, F(2, 473) = 10.34, p < .01. Specifically, post hoc analyses (Tukey's LSD) indicated that adolescents (aged 13 to 18) reported significantly more negative life events than preadolescents (aged 10 to 12) and children (aged 7 to 9, p < .05). Similarly, parents reported differences in their child's negative life events across their child's age, F(2, 473) = 5.35, p < .01. Post hoc analyses again revealed that parents reported significantly more negative life events for adolescents than for preadolescents and children (p < .05). However, SES did not significantly influence reporting of negative life events, F(4, 950) = 2.07, ns.
Additionally, two independent-sample t tests were conducted to determine whether gender had a significant effect on parent and child report of negative life events. No differences were found in child reports of negative life events, t(478) = 1.45, ns. Similarly, parents did not report different numbers of negative life events based on gender, t(477) = 0.55, ns. Fourth, an independent-sample t test was conducted to determine if the gender of the parent had a significant effect on parent reporting of child negative life events. No differences were found in parent report of negative life events, t(477) = 0.91, ns. Finally, a correlation was conducted to determine whether time since diagnosis was correlated with the number of events reported. Time since diagnosis was not related to the total number of events reported, r = –.03, ns.
Based on the above findings, the child's age and race/ethnicity were used as covariates for all subsequent analyses conducted on children's reports of negative life events.
Parent-Child Agreement on the Number of Life Events
Dependent-sample t tests were conducted to compare the total
number of negative life events reported by the child and his or her parent for
the past year and for the lifetime of the child. A statistically significant
difference was found for the number of events reported for the past year,
t(478) = 5.36, p < .001, and for the child's lifetime,
t(478) = 3.26, p < .01, with children reporting more
events in each of these instances. The mean number of events in the past year
reported was 2.04 (SD = 1.89) for children's reports and 1.58
(SD = 1.82) for parents' report. The mean number of events in the
lifetime of the child reported was 5.68 (SD = 3.32) for children's
reports and 5.31 (SD = 3.39) for parents' report.
Number of Reported Life Events in Children's Groups
A multivariate analysis of covariance (MANCOVA) was conducted to compare
the total number of events reported by children and parents in the past year
and in the lifetime of the child. With children's group (cancer, chronically
ill, and healthy) as the independent variable, the omnibus MANCOVA was
significant, F(8, 932) = 2.,73, p < .01, multivariate
h2 = .023. Specific significant differences were found for
children's report of negative life events occurring in the past year,
F(2, 469) = 3.15, p < .05, partial h2
= .013, but not for children's report of occurrence in the child's lifetime,
F(2, 469) = 2.07, ns. Next, significant differences were
found for parents' reports of their child experiencing negative life events
occurring in the past year, F(2, 469) = 6.60, p < .01,
partial h2 = .027, and for parents' reports of negative
life events occurring in the child's lifetime, F(2, 469) = 4.28,
p < .05, partial h2 = .018.
Post hoc analyses (Tukey's LSD) indicated that children in the chronically ill group reported significantly fewer negative life events occurring in the past year when compared with the healthy and cancer groups. However, children in the healthy and cancer groups reported statistically the same amount of negative life events for the past year (p > .05). In contrast, parents with children in the cancer group reported significantly more negative life events for their child in the past year than parents with children in the healthy and chronically ill group. Furthermore, parents with children in the cancer group reported their child as experiencing significantly more negative life events in the child's lifetime than children with a chronic illness (p < .05). Means and standard deviations are listed in Table III.
|
Parent and Child Report Correlations
Significant Pearson product-moment correlations were found for parent and
child report of the child's negative life events in the last year,
r(480) = .49, p < .01, and in the child's lifetime,
r(480) = .72, p < .01. A post hoc r to
z transformation indicated that these correlations were statistically
different (zobt = 14.92, p < .01), indicating
that the correlation between parent and child reported life events was
significantly stronger than for events in the past year.
Percent Agreement
A MANCOVA was conducted comparing the three groups of children (cancer,
chronically ill, and healthy) on the percent agreement of reported negative
life events for the past year and for the lifetime of the child. Percent
agreement was calculated by dividing the number of items agreed upon by the
parent and the child by the total number of events (items agreed / total
number of items). Percent agreement was examined in addition to the total
number of life events in order to obtain a fine-grained comparison of child
and parent report.
The omnibus MANCOVA was significant, F(4, 938) = 7.87, p < .001, partial h2 = .032. Univariate analyses indicated that there was a significant difference between children's groups for percent agreement of negative life events in the last year, F(1, 470) = 36.62, p < .001, partial h2 = .072; and in the lifetime of the child, F(1, 470) = 17.44, p < .001, partial h2 = .036. Post hoc analyses indicated that healthy children and their parents exhibited significantly lower percent agreement in their reports for the child's lifetime than the cancer and chronically ill groups. Regarding the past year, parents of healthy children exhibited higher agreement than parents of children with a chronic illness. Additionally, in order to determine parent and child agreement based on age, a one-way MANOVA was conducted comparing children, preadolescents, and adolescents. No differences in percent agreement were found based on the age of the child, F(4, 946) = 1.87, ns. Means and standard deviations are listed in Table IV.
|
To account for chance in percent agreement, Kappa statistics for each item were analyzed comparing child and parent report. Basically, Kappa is an index that compares agreement between reporters against that which might be expected by chance alone. Significant Kappas were found for all question items regardless of children's group. This finding indicates that child and parent report of negative life events are relatively consistent. A mean Kappa was calculated for each group to more effectively describe agreement across groups.
For lifetime report of negative events, the mean Kappas were .72 for the cancer group, .60 for the chronically ill group, and .54 for the healthy group. An ANOVA revealed that the mean Kappa for lifetime report of events differed by group, F(2, 55) = 5.84, p < .01, partial h2 = .18. Similarly, for events reported as occurring in the last year, the mean Kappas were .62 for the cancer group, .48 for the chronically ill group, and .41 for the healthy group. An ANOVA revealed that the mean Kappa for events reported in the past year differed by group, F(2, 61) = 6.51, p < .01, partial h2 = .18. Specifically, post hoc analyses (Tukey's LSD) revealed that the cancer group had a higher Kappa than the chronically ill group and the healthy group (p < .05) for events reported in both the past year and the lifetime of the child. Although these results should be interpreted with caution, they suggest that the cancer group differed with regard to parent-child agreement in reporting negative life events. Individual item and group mean Kappas for the three children's groups are listed in Table II.
Additionally, three dependent-sample t tests, based on children's group, were conducted to examine total child report of life events and total parent report of life events. Overall, children in the healthy group reported significantly more negative life events than their parents reported for them, t(279) = 3.35, p < .01, but children and parents in the cancer and chronically ill groups did not report statistically differing amounts of negative life events, t(114) = –.39, ns, and t(83) = 1.76, ns, respectively. Means and standard deviations are reported in Table III.
|
Discussion |
|---|
|
TopAbstractIntroductionDiscrepancy in Parent and...Differences in Reporting of...Goals of the StudyMethodsResultsDiscussionReferences |
|---|
Negative life events, whether reported by children or their parents, have been associated with negative developmental outcomes and consequences (Gersten et al., 1977
;
Heisel et al., 1973;
Hodges et al., 1990;
MacLean et al., 1992;
Sandler & Ramsey, 1980).
Unfortunately, discrepancies have been demonstrated between parents and
children regarding the occurrence of the events
(Bailey & Garralda, 1990;
Rende & Plomin, 1991;
Sandberg et al., 2001), and
such disagreement may impact the quality of care that children receive in
response to the negative event (Yeh &
Weisz, 2001). The goals of the present investigation were (1) to
examine parent-child agreement on one measure of negative life events in
healthy and pediatric samples, (2) to examine differences in the frequency of
negative life events across pediatric illness conditions, and (3) to examine
parent-child agreement across specific time frames.
Previous investigations have demonstrated that children report
significantly more negative life events than their parents report for them
(e.g., Bailey & Garralda,
1990). Consistent with these results, we found that children
reported more negative life events than their parents reported. However, to
extend the literature, we examined these data from a number of different
standpoints. Although the absolute number of events reported by children and
parents differed, we also examined the correlation between number of events
and found significant positive correlations between child and parent reports
both for number of events reported for the past year and for events reported
over the child's lifetime.
As a more conservative (and more clinically relevant) examination of agreement, we calculated Kappa for specific negative life events: Whether parents and child agree on a total number of events may be of less clinical importance than whether they agree on specific life events. Further, our use of Kappa represents a methodological advantage in that chance agreement between respondents is taken into account when determining significance. As indicated in Table IV, most items yielded relatively high Kappa coefficients, suggesting that even fairly young children (e.g., age 7) appear able to provide reasonably consistent responses with their parents on measures of life events.
Anecdotally, we noted that several items exhibited particularly low
agreement. For example, agreement in reporting of the death of a relative or
close friend was particularly low. Additionally, items such as "parents
spending less time at home" and "parents fighting more" also
demonstrated low agreement across reporters. Finally, reporting on items
regarding the serious illness of a family member (e.g., parent, sibling) also
exhibited low agreement between parent and child report. These findings were
consistent across groups and time periods. Because these items may represent
events that are particularly distressing to children, the possibility of
defensive responding on the part of children should not be ruled out (see
Phipps & Steele, 2002).
However, since child-reported distress was not evaluated, this explanation
remains speculative. Future research on specific items that exhibit lower
percent agreement is suggested.
A second goal of our study was to examine child- and parent-reported
negative life events across three groups representing healthy children,
children with lifelong medical conditions (i.e., CF, JRA, and DM), and
children undergoing treatment for a serious but relatively time-limited
illness (i.e., cancer). Previous investigations have examined reports of
negative life events among some of these groups, but no investigations, to
date, have directly compared all three groups. Results were partially
consistent with our hypotheses: Children with a chronic illness reported fewer
negative life events in the past year than healthy children or children with
cancer. Unexpectedly, no significant differences were obtained between
children with cancer and healthy children for child-reported events in the
past year or the lifetime of the child. Consistent with Jacobs and Charles
(1980), however, parents of
children with cancer reported more negative life events occurring in the past
year than parents of healthy children or parents of children with long-term
illnesses.
To address the obvious inconsistency in the reports of absolute numbers of
events reported by children and parents in each children's group, we also
examined parent-child agreement in each group. After accounting for chance
agreement, we found that children with cancer and their parents were the most
consistent in their reporting. Unexpectedly, children with a chronic illness
and children who were healthy did not differ in the level of parent and child
agreement. The finding that children with cancer and their parents were most
consistent in their reporting is difficult to interpret. On the one hand,
higher agreement with parents may suggest better communication between parents
and children, which may be considered a protective factor for children's
mental health (Whelan & Kirkby,
2000). On the other hand, increased agreement may suggest fewer
available coping strategies to defend against the distress associated with
these life events (cf. depressive realism;
Vazquez, 1987). The current
literature does not allow conclusions regarding the relative impact of
negative events (i.e., as would be assessed through parent report or actuarial
data) versus child's perceptions of the event (i.e., as would be assessed
through child self-report). Thus, the psychosocial correlates of negative life
events as well as children's knowledge of negative life events deserve further
evaluation.
In light of the relatively modest effect sizes exhibited when the total number of life events was examined across illness groups, the veracity of these results may be questioned. However, a unique strength of this study was that chance agreement was accounted for by calculating Kappa for each item. Unlike the findings that did not account for chance agreement (e.g., comparison of absolute numbers of events), analyses using Kappa to examine differences across groups produced relatively robust findings (partial h2 = .18) for both life events and past-year events. Based on this difference, we suggest that future investigations employ statistical means to control for chance agreement (e.g., Kappa) when investigating differences in parent- and child-reported life events. Previous investigations have examined only correlations and mean numbers of events reported; we feel that our results are an important contribution to this literature.
Finally, we examined parent-child agreement for negative life events across
two distinct time frames (i.e., the past year and over the child's lifetime).
Consistent with our hypothesis, a stronger relationship was found between
parent and child report for events occurring in the child's lifetime as
compared with events occurring in the past year. This finding is particularly
interesting because children may have limited access to events that occurred
in their preverbal ages. Because children should have limited access to these
events, we would expect less agreement between the parent and the child for
life events. However, this was not the case. Overall, a likely interpretation
of this finding is that parents and children may agree that an event occurred
but disagree on the specific time of occurrence (e.g., 10 vs. 14 months ago).
This view is consistent with literature suggesting that children have a
decreased ability to correctly recall information in a temporal framework
(McCormack & Hoerl, 1999),
but not necessarily decreased ability to recall the event itself. Previous
investigations have not been able to make this distinction when reporting
disagreement for life events.
Two specific clinical implications are noted in light of the present findings. First, clinicians are urged to recognize the strengths and limitations of multiple reporters of negative life events. Consistent with the literature, our results suggest that children report significantly more life events than parents report for them. Our results indicate that this is true across illness categories, as well as across time frames. The relative merit of examining the perception of negative life events (i.e., self-report) versus objective (e.g., actuarial) data regarding the occurrence of the event is left for future researchers to address.
Second, our results suggest that the use of a chronological time frame in the assessment of negative life events may reduce parent-child agreement for the events. We see this as a reflection of children's immature cognitive capacity to appreciate events in the context of time. Thus, we suggest that assessments query the event rather than the event during a specific time frame. However, if information on events within a specific time frame is essential, clinicians are encouraged to use concrete time prompts (e.g., calendars) or clear temporal benchmarks (e.g., "Did ________ happen before your last birthday?") in these instances.
The results from this investigation are somewhat mixed: Children with
long-term chronic illnesses reported fewer negative life events than children
with cancer or healthy children, while parents of children with cancer
reported more negative life events than parents of children in the two other
children's groups. Further, children with cancer appear to report negative
life events in a manner more consistent with their parents. Although this
study replicates some findings (Bailey
& Garralda, 1990; Jacobs
& Charles, 1980), future research is needed to better
understand the differences in reporting between children and their parents.
Specifically, the literature on life events is much broader than that
concerning only negative life events, which were the sole focus of the current
investigation. Positive life events also provide valuable information
regarding the child's overall quality of life, and represent an area toward
which future research should be directed. Additionally, the current study
focused on the number of events and the agreement on whether these events
occurred. Future research examining the reported impact of the events may
provide valuable information as well. Finally, research on the psychosocial
correlates or consequences of parent-child agreement for life events is also
encouraged. Specifically, the relative impact of communication about the event
vis-à-vis the impact of the event itself is unknown. This distinction
may have significant implications regarding children's coping strategies and
resilience.
|
Acknowledgments |
|---|
This study was supported in part by the American Lebanese Syrian Associated Charities (ALSAC), Memphis, Tennessee, and by the National Cancer Institute, Cancer Center Support (CORE) grants CA 21765 and CA 23099.
Received July 19, 2002; revision received January 22, 2003; revision received April 8, 2003; accepted April 10, 2003
|
References |
|---|
|
TopAbstractIntroductionDiscrepancy in Parent and...Differences in Reporting of...Goals of the StudyMethodsResultsDiscussionReferences |
|---|
Bailey, D., & Garralda, M. E. (1990). Life events: Children's reports. Social Psychiatry and Psychiatric Epidemiology, 25, 283 –288.[CrossRef][Web of Science][Medline]
Coddington, R. D. (1972). The significance of life events as etiologic factors in the diseases of children. II. A study of a normal population. Journal of Psychosomatic Research, 16, 205 –213.[CrossRef][Web of Science][Medline]
Dinges, N. G., & Duong-Tran, Q. (1992). Stressful life events and co-occurring depression, substance abuse and suicidality among American Indian and Alaska Native adolescents. Culture, Medicine and Psychiatry, 16, 487 –502.
Forsen, A. (1991). Psychosocial stress as a risk for breast cancer. Psychotherapy and Psychosomatics, 55, 176 –185.[Web of Science][Medline]
Gersten, J. C., Langner, T. S., Eisenberg, J. G., & Simcha-Fagan, O. (1977). An evaluation of the etiologic role of stressful life-change events in psychological disorders. Journal of Health and Social Behavior, 18, 228 –244.[CrossRef][Web of Science][Medline]
Heisel, J. S., Ream, S., Raitz, R., Rappaport, M., & Coddington, R. D. (1973). The significance of life events as contributing factors in the diseases of children. 3. A study of pediatric patients. Journal of Pediatrics, 83, 119 –123.[CrossRef][Web of Science][Medline]
Hodges, W. F., London, J., & Colwell, J. B. (1990). Stress in parents and late elementary age children in divorced and intact families and child adjustment. Journal of Divorce and Remarriage, 14, 63 –79.
Hollingshead, A. B. (1975). Four factor index of social status. Unpublished ms., Yale University, Department of Sociology.
Jackson, Y., & Warren, J. S. (2000). Appraisal, social support, and life events: Predicting outcome behavior in school-age children. Child Development, 71, 1441 –1457.[CrossRef][Web of Science][Medline]
Jacobs, T. J., & Charles, E. (1980). Life events
and the occurrence of cancer in children. Psychosomatic
Medicine, 42, 11
–24.
Johnson, J. H., & McCutcheon, S. M. (1980). Assessing life stress in older children and adolescents: Preliminary findings with the Life Events Checklist. In I. G. Sarason & C. D. Spielberger (Eds.), Stress and anxiety (Vol. 7, pp. 111 –125). Washington, DC: Hemisphere.
Kager, V. A., & Holden, E. W. (1992). Preliminary
investigation of the direct and moderating effects of family and individual
variables on the adjustment of children and adolescents with diabetes.
Journal of Pediatric Psychology,
17, 491
–502.
MacLean, W. E., Perrin, J. M., Gortmaker, S., & Pierre, C. B.
(1992). Psychological adjustment of children with asthma: Effects
of illness severity and recent stressful life events. Journal of
Pediatric Psychology, 17, 159
–171.
Mandal, J. M., Ghosh, R., & Nair, L. (1992). Early childhood experiences and life events of male cancer patients, psychosomatic patients and normal persons: A comparative study. Social Science International, 8, 44 –49.
McCormack, T., & Hoerl, C. (1999). Memory and temporal perspective: The role of temporal frameworks in memory development. Developmental Review, 19, 154 –182.[CrossRef]
Phipps, S., & Steele, R. G. (2002). Repressive
adaptive style in children with chronic illness. Psychosomatic
Medicine, 64, 34
–42.
Rende, R. D., & Plomin, R. (1991). Child and parent perceptions of the upsettingness of major life events. Journal of Child Psychology and Psychiatry and Allied Disciplines, 32, 627 –633.
Sandberg, S., Rutter, M., Pickles, A., McGuinness, D., & Angold, A. (2001). Do high-threat life events really provoke the onset of psychiatric disorder in children? Journal of Child Psychology and Psychiatry and Allied Disciplines, 42, 523 –532.[CrossRef]
Sandler, I. N., & Ramsey, T. B. (1980). Dimensional analysis of children's stressful life events. American Journal of Community Psychology, 8, 285 –302.[CrossRef][Web of Science][Medline]
Swearingen, E. M., & Cohen, L. H. (1985). Life events and psychological distress: A prospective study of young adolescents. Developmental Psychology, 21, 1045 –1054.[CrossRef]
Vazquez, C. (1987). Judgment of contingency: Cognitive biases in depressed and nondepressed subjects. Journal of Personality and Social Psychology, 52, 419 –431.[CrossRef][Web of Science][Medline]
Vinnick, L. A., & Erickson, M. T. (1992). Relationships among accumulated lifetime life events, prosocial skills, and behavior problems in elementary school children. Journal of Child and Family Studies, 1, 141 –154.
Whelan, T. A., & Kirkby, R. J. (2000). Parent adjustment to a child's hospitalisation. Journal of Family Studies, 6, 46 –64.
Yeh, M., & Weisz, J. R. (2001). Why are we here at the clinic? Parent-child (dis)agreement on referral problems at outpatient treatment entry. Journal of Consulting and Clinical Psychology, 69, 1018 –1025.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  K. D. Tessner, V. Mittal, and E. F. Walker Longitudinal Study of Stressful Life Events and Daily Stressors Among Adolescents at High Risk for Psychotic Disorders Schizophr Bull, September 4, 2009; (2009) sbp087v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. W. Varni, C. Limbers, and T. M. Burwinkle Literature Review: Health-related Quality of Life Measurement in Pediatric Oncology: Hearing the Voices of the Children J. Pediatr. Psychol., October 1, 2007; 32(9): 1151 - 1163. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. T. Brown, M. Connelly, C. Rittle, and B. Clouse A Longitudinal Examination Predicting Emergency Room Use in Children with Sickle Cell Disease and Their Caregivers J. Pediatr. Psychol., March 1, 2006; 31(2): 163 - 173. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. G. Steele, M. L. Dreyer, and S. Phipps Patterns of Maternal Distress Among Children With Cancer and Their Association With Child Emotional and Somatic Distress J. Pediatr. Psychol., October 1, 2004; 29(7): 507 - 517. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||