Journal of Pediatric Psychology, Vol. 27, No. 4, 2002, pp. 351-361
© 2002 Society of Pediatric Psychology
Validation of a Screening Instrument for Exposure to Violence in African American Children
Louisiana State University
All correspondence should be sent to Anise Flowers, who is now at the Tarnow Center for Self-Management, 1001 West Loop South, #215, Houston, Texas 77027. E-mail: doctorflowers{at}email.com .
 |
Abstract |
|---|
 Top Abstract IntroductionMethodResultsDiscussionReferences |
|---|
Objective: To provide concurrent validity data for the KID-SAVE as a screening instrument for exposure to violence in African American children, to explore demographic differences in KID-SAVE scores, and to provide preliminary reliability data on a parent version of the KID-SAVE.
Method: Questionnaire data were collected regarding exposure to violence, children's behavior and symptoms, and family aggression. A sample of 182 children and their parents participated.
Results: Both parent and child report of violence exposure was significantly related to the child's psychological adjustment as endorsed by both parents and children. Also, significant relationships were obtained between parent report of their child's exposure to violence and the presence of family violence.
Conclusions: The KID-SAVE appears to be a promising instrument for the assessment of exposure to violence, specifically in African American children, and may be applicable in a variety of clinical settings.
Key words: violence exposure; assessment; minority children.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
Exposure to community violence is a growing public health concern for children in the United States. In addition to high rates of community violence in urban cities, several highly publicized episodes of school violence have occurred in rural communities during recent years. In 1995, 35% of homicide victims in the United States were under 24 years of age. The rate of homicide for black males under age 14 was 69% higher than the rate for white males (Centers for Disease Control and Prevention [CDC], 1996
). Each day in the United States, 11 children die from
a gunshot wound, and homicide is the second leading cause of death among
adolescents (Children's Defense Fund
Report, 1996). Estimates of the incidence of elementary school
children witnessing a shooting or stabbing has ranged from 18% to 30%
(Bell & Jenkins, 1993;
Osofsky, Wewers, Hann, & Fick,
1993; Richters & Martinez,
1993b). Even more alarming, one study found that 79% of African
American children had witnessed a shooting and 95% had heard gunshots
(Jones, Ajirotutu, & Johnson,
1996).
Exposure to violence may result in a variety of difficulties, including
posttraumatic stress disorder or other symptoms such as depression, anxiety,
hopelessness, aggression, poor academic achievement, and
internalizing/externalizing behavior difficulties
(Horn & Trickett, 1998;
Osofsky, 1995;
Schwab-Stone et al., 1995;
Singer, Anglin, Song, & Lunghofer,
1995). Children may be exposed to different types of violence,
such as family violence, physical/sexual abuse, or community violence. The
research on community violence focuses on the child's neighborhood and school,
beyond events occurring within the home. Generally, studies on the effects of
community violence exposure have used child self-report measures. Overall,
children's report of their exposure correlates significantly with their report
of related symptoms. However, some studies have found parents often
underreport their child's violence exposure, and other studies indicate no
relationship between parent and child report
(Hill & Jones, 1997;
Horn & Trickett, 1998;
Martinez & Richters,
1993).
Inconsistent findings have been reported regarding age and gender
differences in children's exposure to community violence. With regard to the
frequency of exposure, several studies have reported no age differences
(Cooley-Quille, Turner, & Beidel,
1995b; Fitzpatrick &
Boldizar, 1993; Richters &
Martinez, 1993b). Some studies have found differences in impact,
with younger children reporting greater depression and worse long-term
recovery than older youth (Cooley-Quille et
al., 1995a; Fitzpatrick,
1993; Fitzpatrick &
Boldizar, 1993; Warner &
Weist, 1996). Several studies have found gender to be a
significant predictor of traumatic symptoms, with females displaying greater
symptomatology (Fitzpatrick,
1993; Fitzpatrick &
Boldizar, 1993; Singer et al.,
1995), while another study found maladaptive outcomes following
violence exposure only in males
(Cooley-Quille et al., 1995b).
Some researchers have failed to find differences in level of exposure across
gender (Bell & Jenkins,
1993; DuRant, Cadenhead,
Pendergrast, Slavens, & Linder, 1994), while other researchers
have found that males are significantly more likely to be a victim or witness
of violence (Fitzpatrick & Boldizar,
1993; Gladstein, Rusonis,
& Heald, 1992; Richters & Martinez, 1993;
Selner-O'Hagan, Kindlon, Buka, Raudenbush,
& Earls; 1998; Singer et
al, 1995).
Ethnic factors should be considered, as ethnic minorities are often at a
higher risk for exposure to community violence. In particular, African
American families often live in neighborhoods fraught with poverty, drug
deals, and violent crime. Much of the research on the effects of community
violence exposure has been conducted with samples containing only African
American children (Horn & Trickett,
1998). Studies have indicated that African American children have
a higher incidence of witnessing violence and being a victim of violence than
their white counterparts (Gladstein et
al., 1992; Hammond & Yung,
1994; Selner-O'Hagan et al.,
1998). Also, perpetrators of violent acts are more often family
members, friends, and acquaintances than strangers committing a random crime
(Hammond & Yung, 1994). An
empirically validated instrument to screen for the incidence of exposure to
community violence may be helpful for identifying African American children in
need of further intervention.
Although the literature has suggested a negative impact of violence
exposure, a more precise understanding of the incidence and impact
necessitates the use of empirically developed instruments. Unfortunately,
research on the effects of violence has been hindered by a lack of
psychometrically sound measures. Several instruments used to assess exposure
to community violence lack adequate psychometric properties
(Attar, Guerra, & Tolan,
1994; Cooley-Quille, Turner,
& Beidel, 1995a;
Fitzpatrick, 1993;
Gladstein et al., 1992;
Richters & Martinez, 1990;
Richters & Martinez,
1993b). Often, researchers have generated and categorized scale
items in an unsystematic, nonempirical manner. Subscales are generated from a
logical rationale, but are not verified empirically. Several scales for
children have subscales that lack empirical support or encompass a relatively
narrow spectrum of violent events
(Cooley-Quille et al., 1995a;
Richters & Saltzman,
1990). In general, measures of violence exposure have had
questionable internal consistency with little data available on test-retest
reliability, validity, or norms (Attar et
al., 1994; DuRant, Getts,
Cadenhead, Emans, & Woods, 1995;
Fitzpatrick & Boldizar,
1993; Richters,
1990; Richters & Saltzman,
1990).
The Screen for Adolescent Violence Exposure (SAVE;
Hastings & Kelley, 1997)
was developed as an adolescent self-report scale for assessing frequency of
violence exposure in settings relevant to adolescent adjustment (home, school,
and neighborhood). Scale items were generated from participants living in
high-crime neighborhoods rather than rationally derived. Reliability and
validity data were obtained from a sample of over 1,200 innercity youths, 94%
of whom were African American. Factor structure, examined by both exploratory
and confirmatory factor analyses, identified three factors for each setting
scale: traumatic violence, indirect violence, and physical/verbal Abuse.
Internal consistency and test-retest reliability estimates were acceptable.
Construct, convergent, and known-groups validity (successful classification
into high or low violence groups based on data from the police, school, and
parents) have also been demonstrated for the SAVE
(Hastings & Kelley,
1997).
The KID-SAVE was developed as a modified version of the SAVE to provide a
valid measure of violence exposure in a younger age group
(Flowers, Hastings, & Kelley,
2000). The KID-SAVE was developed on 470 children in grades 3
through 7 who lived in an inner-city area, and the sample was 90% African
American. The KID-SAVE was examined for factor structure using a principle
component analysis with a varimax rotation. The most meaningful factor
structure was obtained by forcing three factors that accounted for 37.2% of
the variance on the Frequency scale. The items clustered according to severity
into three groups: (1) Traumatic Violence: witnessing a shooting or murder, or
being the victim of an assault with a deadly weapon (12 items); (2) Indirect
Violence: witnessing less severe interpersonal violence or hearing about
violent events (16 items); (3) Physical/Verbal Abuse: hitting among peers and
grown-ups hitting/screaming at child (6 items).
In the development study, internal consistency estimates range from.60
to.91 and temporal reliability estimates range from.58 to.86 for Frequency and
Impact scales. Initial validity data were obtained through comparisons between
the KID-SAVE and the Trauma Symptom Checklist for Children (TSCC;
Briere, 1996). Five subscale
scores of the TSCC were significantly correlated with all six KID-SAVE
Frequency and Impact subscale scores, indicating a positive relationship
between reports of violence exposure and symptoms associated with traumatic
stress. One limitation of the validity analyses was method invariance, as TSCC
scores and KID-SAVE scores were both obtained through means of
self-report.
The KID-SAVE appears to be a promising instrument for assessing the frequency and impact of violence exposure in school-age children and has acceptable psychometric properties. The general purpose of this study was to further evaluate the validity of the KID-SAVE as a screening instrument for African American children. The first goal of this study was to examine the concordance between parent and child report of violence exposure incidence and impact. Second, age and gender differences in the rate of exposure to violence reported by the children were evaluated. Based on previous research, girls and younger children were expected to have higher impact scores, whereas boys would have higher frequency scores. Third, relationships between the violence exposure, reported by both children and parents, and behavior problems or psychological symptoms were explored. Finally, the relationship between family violence reported by parents and their child's violence exposure was investigated.
|
Method |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
Participants
A sample of 182 children and their parents was recruited to participate in the study. Families were recruited to participate in the study through the public school system (64% of sample) or from the pediatric clinic of a large urban hospital (36%). Participating children attended public schools that primarily serve youths from high-crime neighborhoods in a large Southern city. The sample was African American and included 94 boys and 88 girls from grades 3 through 7, with an average age of 9.88 years (SD = 1.65). Fifty-eight percent of the families reported an annual family income under $20,000, and a total of 78% reported an income less than $30,000. Nearly all of the parents approached in the pediatric clinic participated in the study. In the schools, the participation rate ranged from approximately 4% to 10%, as participation in this setting required parents to come to the school and stay late with their child after the school day had ended.
Measures
KID-SAVE. The KID-SAVE
(Flowers et al., 2000) is a
measure of violence exposure for children in grades 3 through 7. The scale
consists of 34 items loading onto three subscales: Traumatic Violence,
Indirect Violence, and Physical/Verbal Abuse. Each item is evaluated along
three levels of frequency (0 = never, 1 = sometimes, 2 = a lot) and three
levels of impact (0 = not at all upsetting, 1 = somewhat upsetting, 2 = very
upsetting). In addition, three faces (smiling, frowning, very upset)
accompanied the three levels of impact to assist children in identifying an
appropriate answer. The scale has demonstrated adequate internal consistency
(Frequency: Indirect Violence =.88, Traumatic Violence =.82, Physical/Verbal
Abuse =.60; Impact: Indirect Violence =.85, Traumatic Violence =.77,
Physical/Verbal Abuse =.62) and temporal reliability
(Flowers et al., 2000). A
parent version of the KID-SAVE containing the same items was administered to
parents. The parent version is simply a rewording of the child version and has
not yet been empirically evaluated.
Trauma Symptom Checklist for Children (TSCC-A). The TSCC-A
(Briere, 1996) is designed to
assess trauma-related psychological symptoms. The TSCC-A is a shortened
version of the original TSCC, which does not include a subscale for sexual
concerns. This alternate version contains 44 items and consists of six
subscales measuring constructs associated with the development of PTSD:
Anxiety, Depression, Posttraumatic Stress, Dissociation, Anger, and two
validity subscales. Internal consistency for symptom subscales ranges from.82
to.89. Acceptable convergent validity has been demonstrated between the TSCC-A
and general measures of child psychopathology. In addition, evidence for
construct validity has been established through relationships between TSCC-A
scores and stressful life events or exposure to violence
(Briere, 1996).
Child Behavior Checklist (CBCL). The CBCL
(Achenbach, 1991) is a 112-item
parent report scale of childhood and adolescent problem behaviors with
subscales for internalizing and externalizing behavior disorders. Subscales
include Withdrawn, Somatic Complaints, Anxious/Depressed, Social Problems,
Thought Problems, Attention Problems, Delinquent Behavior, Aggressive
Behavior, Externalizing Behaviors, Internalizing Behaviors, and Total
Problems. Reliability and validity for the CBCL has been well established in
the child psychopathology literature. Internal consistency estimates for the
subscales ranges from.68 to.92 with an alpha of.96 for Total Problems
(Achenbach, 1991,
1998).
Revised Conflict Tactics Scale (CTS2). The CTS2
(Straus, Hamby, Boney-McCoy, &
Sugarman, 1996) is a measure of how couples resolve conflicts with
subscales for Negotiation, Psychological Aggression, Physical Assault, and
Injury. Each of the 32 items is rated for frequency of occurrence in the
previous year for the respondent and their partner. Internal consistency
estimates range from.79 to.95 and initial evidence exists for construct
validity (Straus et al.,
1996). The CTS2 may provide an index of violence level in the
home.
Procedure
After obtaining informed consent from a parent or guardian and assent from
the child; the battery of questionnaires, including a demographic sheet, was
administered the child and parent. The institutional review boards of the
participating hospital and university granted approval for the study. The
children completed the KID-SAVE and TSCC-A, and parents completed the
P-KID-SAVE, CBCL, and CTS2. Participants were informed that the researcher was
interested in obtaining anonymous information regarding the child's experience
with violent events and related feelings. Instructions for completing the
measures were reviewed prior to administration, and all questionnaires were
read aloud to elementary-age students to ensure proper understanding of the
items. In the hospital setting, questionnaires were individually administered.
In the schools, data collection took place after school and questionnaires
were administered to parents and children as a group. Each family received $15
for their participation in the study.
|
Results |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
Results are presented regarding the reliability of the parent version of the KID-SAVE and the agreement between parent and child report on questionnaires of exposure to violence. Age and gender differences in children's report of violence exposure were explored using MANOVA. Multiple regression analyses are presented regarding the relationships between violence exposure and symptoms reported by children and parents. Finally, MANOVA analyses were conducted to explore the relationship between children's violence exposure and family violence.
Samples
No significant differences were found between subjects recruited from
schools versus those from the hospital on t test results for subscale
scores of the KID-SAVE, P-KID-SAVE, CBCL, TSCC-A, or the relevant subscale of
the CTS2. The two sample groups were combined for all further analyses.
Reliability Analyses
Reliability was investigated for the subscales of both the parent and child
versions of the KID-SAVE. Cronbach's alpha was calculated for both the
Frequency and the Impact subscales. For the child version, reliability results
for the Frequency subscales were Indirect Violence =.82, Traumatic Violence
=.74, Physical/Verbal Abuse =.68; and for the Impact subscales: Indirect
Violence =.81, Traumatic Violence =.66, Physical/Verbal Abuse =.64. The
estimates for several subscales were slightly lower than those obtained in the
original sample with a larger sample size. On the parent version, internal
consistency estimates for the Frequency subscales were Indirect Violence =.88,
Traumatic Violence =.65, Physical/Verbal Abuse =.66; and for the Impact
subscales: Indirect Violence =.87, Traumatic Violence =.79, Physical/Verbal
Abuse =.69. Overall, reliability analyses for the parent version indicate that
the measure is internally consistent, and reliability estimates are similar to
those previously obtained for the child version
(Flowers et al., 2000).
Parent-Child Concordance: KID-SAVE
Correlational analyses were performed to examine the concordance between
parent and child reports of violence exposure on the KID-SAVE. With the
exception of the Indirect Violence Impact subscale, the remaining five
subscales were significantly correlated across the child and parent versions
of the KID-SAVE (see Table I). A paired-samples t test was conducted to evaluate whether there were
significant differences in subscale scores between the parent and child
versions of the KID-SAVE. Using the Bonferroni method, each t test
was tested at the.008 level. The results indicated a significant difference
only on the Frequency Physical/Verbal Abuse subscale (t [177] = 2.91,
p <.005). Children reported a significantly higher incidence of
events on the Physical/Verbal Abuse subscale. Parents and children did not
indicate different rates of violence exposure for the Traumatic or Indirect
Violence subscales. Subscale means and standard deviations for parent and
child versions can be found in Table
II.
|
|
Demographic Differences on KID-SAVE Scores
KID-SAVE scores from the child version were examined for grade and gender
differences on the Frequency and Impact subscale scores. A one-way
multivariate analysis of variance was conducted to evaluate the relationship
between gender and the level of violence exposure reported by the child. The
dependent variables included the three Frequency and three Impact subscales
from the KID-SAVE. Significant differences were found between boys and girls
on the dependent measures (Wilks' lambda =.93), F(6, 176) = 2.15,
p <.05. Analyses of variance on each dependent variable were
conducted as follow-up tests to the MANOVA. The ANOVA on the Frequency
Physical/Verbal Abuse subscale was significant, F(1, 181) = 10.7,
p <.005, with boys scoring higher than girls. This gender
difference suggests that boys are experiencing higher rates of interpersonal
fighting and aggression than girls.
Children were combined into groups based on whether they attended an elementary or middle school for the purposes of analyzing age differences. A one-way multivariate analysis of variance was conducted to evaluate the relationship between grade and the level of violence exposure reported by the child. The dependent variables included the three Frequency and three Impact subscales from the KID-SAVE. Significant differences were found between elementary and middle school children on the dependent measures (Wilks' lambda =.78), F(6, 172) = 7.73, p <.001. Analyses of variances on each dependent variable were conducted as follow-up tests to the MANOVA. The ANOVA was significant for the Frequency Indirect Violence subscale, F(1, 177) = 9.8, p <.005, with middle school students scoring higher than elementary age students. The ANOVA also was significant for the Frequency Physical/Verbal Abuse subscale, F(1, 177) = 8.8, p <.005, and the Impact Physical/Verbal Abuse subscale, F(1, 177) = 8.1, p <.01. For both Physical/Verbal Abuse scales, the elementary school students scored significantly higher than middle school students. These age differences indicate that middle school students are experiencing increased rates of witnessing or hearing about serious violent events, while elementary age students have higher rates of interpersonal aggressive acts and are more distressed by these events.
Violence Exposure and Associated Symptoms
Twelve regression analyses were conducted using the Frequency subscales
scores of the KID-SAVE and the Parent-KID-SAVE to predict symptoms. Using the
Bonferroni method, each model was tested at the.004 level. Subscales were
entered into the regression using a forward method in order to determine
specific subscales of the violence measures related to symptoms. For the CBCL,
the Thought Problems subscale was excluded from analyses due to its lower
internal consistency and less clear content validity
(Achenbach, 1991). The Parent
Frequency Physical/Verbal Abuse subscale was significantly related to the
Withdrawn score (R2 =.07, adjusted R2
=.06), F(1, 159) = 11.06, p <.004. The parent
Physical/Verbal Abuse and the child Traumatic Violence subscales emerged as
significant predictors of the Anxious/Depressed score (R2
=.15, adjusted R2 =.14), F(2, 158) = 13.98,
p <.001. The parent Physical/Verbal Abuse, parent Traumatic
Violence, and child Indirect Violence subscales emerged as significant
predictors of the Social Problems score (R2 =.20, adjusted
R2 =.19), F(3, 160) = 13.29, p <.001.
The parent Physical/Verbal Abuse subscale emerged as a significant predictor
of the Attention Problems score (R2 =.06, adjusted
R2 =.06), F(1, 160) = 10.94, p <.004.
The child Traumatic Violence and the parent Physical/Verbal Abuse subscales
emerged as significant predictors of the Delinquent score
(R2 =.09, adjusted R2 =.08),
F(2, 160) = 8.17, p <.001. The parent Physical/Verbal
Abuse and the child Traumatic Violence subscales emerged as a significant
predictors of the Aggression score (R2 =.08, adjusted
R2 =.06), F(2, 160) = 6.50, p <.004.
None of the frequency violence subscales was a significant predictor of the
Somatic score. Overall, parent report of events on the Physical/Verbal Abuse
subscale was predictive of both internalizing and externalizing symptom
scales. In addition, children reporting higher rates of exposure for the
Indirect Violence subscale had parents indicating they have more social
problems. Parents also indicated more difficulty with anxious/depressed,
aggressive, and delinquent behaviors when their children were reporting a
higher frequency of exposure to very serious, traumatic acts of violence.
Similar regression analyses were conducted using subscales of the TSCC-A. Anxiety scores of the TSCC-A were significantly associated with the following violence subscales: child Indirect Violence, child Physical/Verbal Abuse, child Traumatic Violence, and parent Indirect Violence (R2 =.26, adjusted R2 =.24), F(4, 168) = 14.85, p <.001. The child Physical/Verbal Abuse and the child Indirect Violence subscales emerged as significant predictors of the Depression subscale (R2 =.27, adjusted R2 =.26), F(2, 170) = 31.69, p <.001. For both the Anger and Posttraumtic Stress subscales, violence scores entered in the following order: child Indirect Violence, child Traumatic Violence, and child Physical/Verbal Abuse. On the overall model for the Anger subscale the following were obtained: (R2 =.35, adjusted R2 =.34), F(3, 169) = 30.72, p <.001, and on the Posttraumatic Stress subscale, results were (R2 =.24, adjusted R2 =.23), F(3, 169) = 18.10, p <.001. The child Indirect Violence and the child Physical/Verbal Abuse subscales emerged as significant predictors of the Dissociative subscale (R2 =.20, adjusted R2 =.20), F(2, 174) = 22.34, p <.001. Thus, children's report of exposure to violence on all three subscales of the KID-SAVE was highly predictive of symptoms on the TSCC-A. Also, parental report of exposure to indirect violence (events such as witnessing or hearing about serious violent events) was positively related to children's level of anxiety symptoms.
Family Violence and Child Violence Exposure
A one-way multivariate analysis of variance was conducted to evaluate the
relationship between physical violence between caregivers and parent's report
of child violence exposure. The independent measure, physical assault, was
developed by grouping cases into those in which any incidence of physical
assault was reported for the caregiver or their partner on the Conflict
Tactics Scale (CTS2). The relevant subscales from the CTS2 assessed instances
of physical assault between the adult caregivers and did not include
aggressive acts directed at the child. The dependent variables included the
three Frequency and three Impact subscales from the Parent-KID-SAVE.
Significant differences were found on the dependent measures (Wilks' lambda
=.91), F(6, 132) = 2.31, p <.05. Analyses of variances on
each dependent variable were conducted as follow-up tests to the MANOVA. The
ANOVA on the Frequency Indirect Violence subscale was significant,
F(1, 133) = 7.0, p <.01, and the Impact Indirect Violence
subscale was significant, F(1, 133) = 8.7, p <.005. For
both subscales, parents who reported physical assault occurring in the home
also reported significantly more indirect exposure to violence for their
children. Table III contains
the means and the standard deviations on the dependent variables for the two
groups. Another one-way multivariate analysis of variance was conducted to
evaluate the relationship between physical violence between caregivers and the
child's report of violence exposure. No significant relationship was found
using child report on the KID-SAVE.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
The purpose of this study was to provide additional psychometric data on the validity of the KID-SAVE and to explore relationships with psychological symptoms. The KID-SAVE is a screening measure for violence exposure for children in the third through seventh grades, containing 34 items rated for event frequency and emotional impact. Previous research has established adequate internal consistency and test-retest reliability and a three factor structure for the KID-SAVE (Flowers et al., 2000
).
The Parent-KID-SAVE, an analogous parent version of the KID-SAVE, was found
to have adequate internal consistency for both Frequency and Impact subscales.
A significant difference between the mean scores of parent and child report
was found only for the Frequency Physical/Verbal Abuse subscale. This subscale
contains items pertaining to less serious, everyday experiences such as being
hit or threatened by a peer, which children may not report to their parents.
Contrary to previous studies in which parents have underestimated their
child's exposure to violence (Hill &
Jones, 1997; Richters &
Martinez, 1993b), parents and children did not differ in their
report of the frequency of the more serious violent events on the Indirect
Violence or Traumatic Violence subscales. Although parents have sometimes
underestimated the emotional impact of exposure to violent events,
(Hill & Jones, 1997;
Schwartz, Gladstone, & Kaslow,
1998) the mean scores for parent and child report on the Impact
subscales were not significantly different in this study. Overall, the parent
and child congruence on five subscales provides cross-informant concurrent
validity for the KID-SAVE.
Previous research regarding children and exposure to violence has provided
inconsistent evidence regarding developmental or gender differences
(Bell & Jenkins, 1993;
Cooley-Quille et al., 1995b;
Fitzpatrick, 1993;
Richters & Martinez,
1993a; Selner-O'Hagan et al.,
1998; Singer et al.,
1995). In this study, boys reported higher scores on the Frequency
Physical/Verbal Abuse subscale than girls. In general, developmental research
indicates that boys are more likely to engage in overt aggressive behaviors or
display externalizing behavior problems
(Campbell, 1998). As might be
expected, boys were reporting more instances of interpersonal aggression on
this subscale of the KID-SAVE. Consistent with some previous studies,
significant gender differences for the more serious violence subscales were
not observed (Bell & Jenkins,
1993; DuRant et al.,
1994). Middle school students had significantly higher Frequency
scores than elementary-age children on the Indirect Violence subscale. On the
other hand, elementary-age children reported significantly higher scores on
the Frequency and Impact of Physical/Verbal Abuse. Although the precise
developmental process is not known, the incidence of events on the
Physical/Verbal Abuse subscale reported by elementary-age students may
decrease as aggressive events become more serious in middle school, similar to
the items on the Indirect Violence subscale. In addition, when middle school
students have a higher rate of experiences like seeing someone carry a gun or
hearing about someone getting shot, one might expect that interpersonal events
such as a peer threatening to beat you would be reported as less
upsetting.
In this study, the report of violence exposure and symptomatology was
investigated across respondents, utilizing both parents and their children.
From a measurement standpoint, significant correlations on several self-report
questionnaires from the same respondent would be expected
(Spector, 1992). Previous
research has found significant relationships between violence exposure and
symptoms only within the same respondent, either parent or child
(Horn & Trickett, 1998).
However, this investigation found a number of significant relationships across
respondents, between parents and children.
Multiple regression analyses were utilized to investigate the relationship
between the parent and child violence scores and symptoms or child behavior
problems. Parental report of the frequency of lower level interpersonal events
from the Physical/Verbal Abuse scale was the most predictive, contributing to
the model for six clinical subscales of the CBCL
(Achenbach, 1991). Several
cross-informant relationships were observed between children's report of
violence exposure and parental report of child behavior problems.
Specifically, higher childreport for the frequency of Indirect Violence was
associated with parental report of social problems. In addition, the frequency
of more serious events on the Traumatic Violence scale as endorsed by children
was predictive of the Anxious/Depressed, Delinquent Behavior, and Aggressive
Behavior subscales of the CBCL. The finding that children who reported
experiencing the most serious traumatic events displayed more behavior or
emotional problems is consistent with the literature.
Children's report of violence exposure was significantly associated with
their own report of symptomatology. This finding is consistent with previous
research on the KID-SAVE and other studies investigating child-report of
violence and symptoms (Fitzpatrick,
1993; Flowers et al.,
2000, Freeman, Mokros, &
Poznanski, 1993; Martinez
& Richters, 1993). Across informants, parents who indicated a
higher level of exposure for Indirect Violence had children who reported
higher levels of anxiety. The significant relationships between both parent
and child data have implications for assessment in clinical settings. In
populations at risk for exposure to community violence, obtaining parent and
child report on a screening measure of violence exposure may be useful in
delineating the role of exposure to violence in a child's current functioning.
Since exposure to violence is related to a variety of internalizing and
externalizing behavior problems (Freeman
et al., 1993; Fitzpatrick,
1993; Fitzpatrick &
Boldizar, 1993; Hill,
Levermore, Twaite, & Jones, 1996;
Warner & Weist, 1996),
obtaining this information may be important for understanding etiology and
differential diagnosis.
Construct validity of the KID-SAVE was evaluated through comparison of
parent report of violence in the home and their child's exposure to violence.
On the Conflict Tactics Scale-2 (Straus et
al., 1996), parents indicated the strategies utilized by self and
partner to resolve conflicts. Parental report of physical assault strategies
for resolving conflict in the home was significantly related to their
reporting of children's exposure to violence. Specifically, parents reported
higher scores for the Frequency and Impact Indirect Violence subscales when
physical assault between caregivers was endorsed. The presence of family
violence is a significant factor in children's overall adjustment and children
have been called the "silent witnesses" as parents sometimes
underestimate both the exposure and the symptoms associated with domestic
violence (Margolin, 1998;
McCloskey, Figueredo, & Koss;
1995; Osofsky &
Scheeringa, 1997; Rhea,
Chafey, Dohner, & Terragno, 1996). Significant concordance has
been documented between being a victim or witness to neighborhood violence and
living in a home with partner violence
(Osofsky et al., 1993).
Although the exact co-occurrence of home and neighborhood violence cannot be
discerned from the KID-SAVE, parents in this sample indicated a connection
between the presence of domestic physical fighting and their child's exposure
to violence. Although a relationship was observed for parent report on the
parent version of the KID-SAVE, a similar relationship was not found for the
children's report. This is a limitation to the construct validity provided by
the parental report of physical aggression in the home.
One limitation of this study is the method invariance encountered through the use of self-report questionnaires. Although several cross-informant relationships were identified, the number was small in comparison to the larger number of significant correlations obtained within respondents. Further validity could be established by relating self-report of violence exposure to objective data regarding home, school, or neighborhood violence statistics. Second, sampling bias may have been present, as participation in the study sometimes required parents to come to their child's school and variability was observed in participation rates. Third, empirical data on the parent version of the KID-SAVE were preliminary and require further investigation. Although the parent version was scored in a similar manner for the purposes of comparison in this study, further data collection and factor analyses may reveal a different subscale composition in the parent version of the KID-SAVE.
The KID-SAVE was developed using African American children who were mainly living in high-crime neighborhoods where exposure to community violence was likely. A screening measure of violence exposure is most needed for use with high-risk children. However, the high-risk nature of the sample limits the generalizibility of the scale to other groups of children. Further validity could be established through comparisons on KID-SAVE scores between high and low exposure samples and investigation with middle-class, Caucasian, or other minority samples.
The KID-SAVE may have clinical utility as a screening instrument to detect
high levels of violence exposure in a community sample of African American
children and identify children who need further assessment and intervention.
Screening children in elementary and middle school would be important, as
research indicates that adolescents who experience significant violence often
do not seek psychological assistance, and school-age children are at even
higher risk for emotional sequelae
(Cooley-Quille et al., 1995b;
Fitzpatrick, 1993;
Hammond & Yung, 1994;
Warner & Weist, 1996).
Among children referred to clinical settings, the KID-SAVE may be utilized
during the assessment process in order to understand the role of violence
exposure in the child's current behavioral or emotional difficulties, as
clinicians may underestimate the child's experiences with witnessing violence
(Guterman & Cameron,
1999).
In summary, the KID-SAVE was empirically developed to provide a screening instrument with research and clinical utility. Both parent and child report of violence exposure was significantly related to the child's psychological adjustment as endorsed by both parents and children. Construct validity has been established through significant relationships between parent report of their child's exposure to violence and the presence of family violence. Overall, the psychometric properties of the KID-SAVE support its utility in assessing the frequency and impact of exposure to violence in young children. The KID-SAVE appears to be a promising instrument for further research regarding the incidence and effects of violence exposure and may be applicable in a variety of clinical settings.
Received January 20, 2001; revision received June 1, 2001; accepted July 15, 2001
|
References |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
Achenbach, T. M. (1991). Manual for the child behavior checklist/4-18 and 1991 profile. Burlington: University of Vermont Department of Psychiatry.
Achenbach, T. M. (1998). Diagnosis, assessment, taxonomy, and case formulations. In T. H. Ollendick & M. Hersen (Eds.), Handbook of child psychopathology. New York: Plenum Press.
Attar, B. K., Guerra, N. G., & Tolan, P. H. (1994). Neighborhood disadvantage, stressful life events, and adjustment in urban elementary-school children. Journal of Clinical Child Psychology, 23, 391-400.[Web of Science]
Bell, C., & Jenkins, E. J. (1993). Community violence and children on Chicago's southside. Psychiatry, 56, 46-54.[Web of Science][Medline]
Briere, J. (1996). Trauma symptom checklist for children: Professional manual. Odessa, FL: Psychological Assessment Resources.
Campbell, S. (1998). Developmental perspectives. In T. H. Ollendick & M. Hersen (Eds.), Handbook of child psychopathology. New York: Plenum Press.
Centers for Disease Control and Prevention. (1996). Homicide deaths and rates per 100,000, United States, 1989-1995. Atlanta, GA: U.S. Dept. of Health and Human Services.
Children's Defense Fund Report. (1996). Children in the United States. Washington, DC: Author.
Cooley-Quille, M. R., Turner, S. M., & Beidel, D.C. (1995a). Assessing community violence: The Children's report of exposure to violence. Journal of the American Academy of Child and Adolescent Psychiatry, 34, 201-215.[Web of Science][Medline]
Cooley-Quille, M. R., Turner, S. M., & Beidel, D.C. (1995b). The emotional impact of children's exposure to community violence: A preliminary study. Journal of the American Academy of Child and Adolescent Psychiatry, 35, 354-369.
DuRant, R. H., Cadenhead, C., Pendergrast, R. A., Slavens, G.,
& Linder, C. W. (1994). Factors associated with the use of
violence among urban black adolescents. American Journal of Public
Health, 84,
612-617.
DuRant, R. H., Getts, A., Cadenhead, C., Emans, S. J., & Woods, E. R. (1995). Exposure to violence and victimization and depression, hopelessness, and purpose in life among adolescents living in and around public housing. Developmental and Behavioral Pediatrics, 16, 233-237.
Fitzpatrick, K. M. (1993). Exposure to violence and presence of depression among low-income, African-American youth. Journal of Consulting and Clinical Psychology, 61, 528-531.[Web of Science][Medline]
Fitzpatrick, K. M., & Boldizar, J. P. (1993). The prevalence and consequences of exposure to violence among African-American youth. Journal of the American Academy of Child and Adolescent Psychiatry, 32, 424-430.[Web of Science][Medline]
Flowers, A. L., Hastings, T. L., & Kelley, M. L. (2000). Development of a screening instrument for exposure to violence in children: The KID-SAVE. Journal of Psychopathology and Behavioral Assessment, 22(1), 91-104.
Freeman, L. N., Mokros, H., & Poznanski, E. (1993). Violent events reported by normal urban school-aged children: Characteristics and depression correlates. Journal of the American Academy of Child and Adolescent Psychiatry, 32, 419-423.[Web of Science][Medline]
Gladstein, J., Rusonis, E. J., & Heald, F. P. (1992). A comparison of inner-city and upper-middle class youths' exposure to violence. Journal of Adolescent Health, 13, 275-280.[Web of Science][Medline]
Guterman, N. B., & Cameron, M. (1999). Young clients' exposure to violence: How much do their therapists know? American Journal of Orthopsychiatry, 69, 382-391.[Web of Science][Medline]
Hammond, W. R., & Yung, B. R. (1994). African Americans. In L. D. Eron, J. H. Gentry, & P. Schlegel (Eds.), Reason to hope: a psychosocial perspective on violence and youth. Washington, DC: APA.
Hastings, T., & Kelley, M. L. (1997). Development and validation of the Screen for Adolescent Violence Exposure (SAVE). Journal of Abnormal Child Psychology, 25, 511-520.[Web of Science][Medline]
Hill, H. J., & Jones, L. P. (1997). Children's and parents' perceptions of children's exposure to violence in urban neighborhoods. Journal of the National Medical Association, 89, 270-276.[Medline]
Hill, H. J., Levermore, M., Twaite, J., & Jones, L. (1996). Exposure to community violence and social support as predictors of anxiety and social and emotional behavior among African American children. Journal of Child and Family Studies, 5, 399-414.
Horn, J. L., & Trickett, P. K. (1998). Community violence and child development: A review of research. In P. K. Trickett & C. J. Schellenbach (Eds.), Violence against children in the family and the community. Washington, DC: American Psychological Association.
Jones, F. C., Ajirotutu, C., & Johnson, J. (1996). African American children and adolescents exposure to community violence: A pilot study. Journal of Cultural Diversity, 3, 48-52.[Medline]
Margolin, G. (1998). Effects of domestic violence on children. In P. K. Trickett & C. J. Schellenbach (Eds.), Violence against children in the family and the community. Washington, DC: American Psychological Association.
Martinez, P., & Richters, J. E. (1993). The NIMH community violence project: II. Children's distress symptoms associated with violence exposure. Psychiatry, 56, 22-35.[Web of Science][Medline]
McCloskey, L. A., Figueredo, A. J., & Koss, M. P. (1995). The effects of systemic family violence on children's mental health. Child Development, 66, 1239-1261.[Web of Science][Medline]
Osofsky, J. D. (1995). The effects of exposure to violence on young children. American Psychologist, 50, 782-788.[Medline]
Osofsky, J. D., & Scheeringa, M. S. (1997). Community and domestic violence exposure: Effects on development and psychopathology. In D. Cicchetti & S. L. Toth (Eds.), Developmental perspectives on trauma: Theory, research, and intervention. New York: University of Rochester Press.
Osofsky, J. D., Wewers, S., Hann, D. M., & Fick, A. C. (1993). Chronic community violence: What is happening to our children? Psychiatry, 56, 36-45.[Web of Science][Medline]
Rhea, M. H., Chafey, K. H., Dohner, V. A., & Terragno, R. (1996). The silent victims of domestic violence—Who will speak? Journal of Child and Adolescent Psychiatric Nursing, 9, 7-15.
Richters, J. E. (1990). Screening survey of exposure to violence. Self report version. Rockville, MD: National Institute of Mental Health.
Richters, J. E., & Martinez, P. (1990). Things I have seen and heard: A structured interview for assessing young children's violence exposure. Rockville, MD: National Institute of Mental Health.
Richters, J. E., & Martinez, P. (1993a). Children as victims of and witnesses to violence in a Washington, D. C. neighborhood. In L. A. Leavitt & N. A. Fox (Eds.), The psychological effects of war and violence on children (pp. 243-278). Hillsdale, NJ: Lawrence Erlbaum.
Richters, J. E., & Martinez, P. (1993b). The NIMH community violence project: I. Children as victims of and witnesses to violence. Psychiatry, 56, 7-21.[Web of Science][Medline]
Richters, J. E., & Saltzman, W. (1990). Survey of Children's Exposure to Community Violence: Parent report. Rockville, MD: National Institute of Mental Health.
Schwab-Stone, M. E., Ayers, T. S., Kasprow, W., Voyce, C., Barone, C., Shriver, T., & Weissberg, R. P. (1995). No safe haven: A study of violence exposure in an urban community. Journal of the American Academy of Child and Adolescent Psychiatry, 34, 1343-1352.[Web of Science][Medline]
Schwartz, J. A. J., Gladstone, T. R. G., & Kaslow, N.J. (1998). Depressive disorders. In T. H. Ollendick & M. Hersen (Eds.), Handbook of child psychopathology. New York: Plenum Press.
Selner-O'Hagan, M. B., Kindlon, D. J., Buka, S. L., Raudenbush, S. W., & Earls, F. J. (1998). Assessing exposure to violence in urban youth. Journal of Child Psychology and Psychiatry, 39, 215-224.[Web of Science][Medline]
Singer, M. I., Anglin, T. M., Song, L., & Lunghofer, L.
(1995). Adolescents' exposure to violence and associated symptoms
of psychological trauma. Journal of the American Medical
Association, 273,
477-482.
Spector, P. E. (1992). Summated rating scale construction: An introduction. Newbury Park, CA: Sage Publications.
Straus, M. A., Hamby, S. L., Boney-McCoy, S., & Sugarman, D. B. (1996). The Revised Conflict Tactics Scale (CTS2). Journal of Family Issues, 17, 283-316.[Abstract]
Warner, B. S., & Weist, M. D. (1996). Urban youth as witnesses to violence: Beginning assessment and treatment efforts. Journal of Youth and Adolescence, 25, 361-377.[Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||