The Clinical and Adaptive Features of Young Offenders with Histories of Child-Parent Violence

Abstract

This study compared the clinical and adaptive features of juvenile offenders (N = 223) who were violent towards their parents (CPV) with those who had no history of violence against their parents (NCPV). These two groups were also examined on demographic data, arrest findings, mental health issues, relationship findings, intellectual abilities, and school performance. Youths in the CPV group were more likely to (a) associate with peers who own guns, (b) affiliate with gang members, (c) belong to a gang, (d) have been psychiatrically hospitalized and medicated, (e) have attempted suicide, (f) come from a non-intact home, and (g) have trouble relating to their parents and other household members. The CPV group also committed a greater number of nondomestic violent offenses, while those in the NCPV group committed a greater number of property offenses. Analyses revealed no significant group differences on the Emotional Symptom Index and Personal Adjustment Composite of the Behavior Assessment System for Children; however, interaction effects were detected by gender and race.

Introduction

Violent crime profoundly impacts the lives of individuals and families across the country. Even though the 2007 arrest rates for Violent Crime Index (VCI) offenses continue to trend lower since the highs of 1994, the data are still alarming (U.S. Department of Justice 2009). The U.S. Department of Justice reported that in 2007 there were 5.2 million U.S. residents age 12 or older who were the victims of violence. Sadly, children and adolescents under the age of 18 accounted for 16% of arrests for these crimes (Federal Bureau of Investigation 2007). Furthermore, it was revealed that family members and friends were frequently the target of juvenile violence. In fact, 52% of victims of violence age 30 and older were either the offender’s parent or stepparent, while 23% of all violent crime victims were family members. More specifically, in crimes committed by juveniles, family members represented 28% of the victims of sexual assault and 24% of the victims of simple assault (U.S. Department of Justice 2009).

Legal and societal concerns about youths who commit crimes have resulted in much research examining the broad domain of juvenile delinquency. Also abundant in the literature are articles on juvenile violence within the home. Child-parent violence (CPV), a subtype of juvenile violence within the home, remains one of the least researched forms of family violence, however the scientific community has recently begun to take a closer look at this destructive phenomenon (Boxer et al. 2009; Edenborough et al. 2008; Kennair and Mellor 2007; Pagani et al. 2004, 2009; Stewart et al. 2007; Ulman and Straus 2003; Walsh and Krienert 2007, 2009). CPV is a serious concern when one considers that disruptive parent-child relationships, of which violence is certainly indicative, can lead to significant distress, subsequent long-term mental and physical difficulties, and potentially devastating consequences for the individual, family and society at large (David et al. 1996; Jaycox and Repetti 1993; Noller 1995; Noller et al. 1992; Robinson et al. 2004; Ulman and Straus 2003; Walsh and Krienert 2009).

Relative to other forms of victimization within the home, CPV may not appear to be as significant a problem; however, considering that the occurrence of such violence is often kept private and only uncovered in unique circumstances, available figures should be viewed as a conservative estimate of the true prevalence of this phenomenon (Gelles 1997; Herron et al. 1994; Jackson 2003). Estimates of CPV have ranged between 5% and 13% (Evans and Warren-Solhberg 1988; Gelles 1997; Paulson et al. 1990; Peek et al. 1985). These ranges have varied in accordance with the degree of violence displayed by the child (Wilson 1996). Most studies published on the topic of CPV have focused on offense-specific characteristics, demographic descriptors, and social, behavioral, and familial correlates (Browne and Hamilton, 1998; Du Bois 1998; Gelles 1997; Micucci 1995; Paulson et al. 1990; Pelletier and Coutu 1992; Robinson et al. 1994; Ulman and Straus 2003; Walsh and Krienert 2009; Wilson 1996). Although important, these findings have provided little information about the psychological functioning of such youths, or how these youths differ from their delinquent counterparts who have no history of CPV. When one considers that assessment techniques and intervention programs are typically built around the clinical needs of the patient, it becomes apparent additional research is warranted.

Child-Parent Violence (CPV)

Harbin and Madden (1979), first identified CPV as a distinct type of family violence. At that time, they reported that almost 10% of youths attacked their parents. Cornell and Gelles (1982), reported that 9% of parents from a nationally representative sample of families reported an act of violence from a youth between the ages of 10 and 17. Peek et al. (1985), in their survey of more than 1,500 parents of White male high school students, revealed that 7%–11% of parents reported being hit by their children.

Findings are mixed as to the relationship between CPV and socioeconomic status (SES). Some researchers found that violence towards parents occurred more often in lower-class families (Paulson et al. 1990; Peek et al. 1985; Wilson 1996), while others suggested that youths who assault their parents were more likely to come from upper–and middle-class families (Agnew and Huguley 1989). When the victim’s gender was examined, it was found that mothers were more likely to be the victims of CPV than fathers (Browne and Hamilton 1998). In fact, the modal pattern of assault was male adolescents against their mothers (Gelles 1997). Another stable pattern to emerge was that males were more apt to be physically abusive while females were more likely to be verbally abusive to their parents (Nock and Kazdin 2002). In addition, even though assaulters were more likely to physically strike their mothers, they were more likely to use excessive force towards their fathers (Browne and Hamilton 1998). Walsh and Krienert (2007) examined a large sample of children and found that the majority (60%–73%) of CPV offenses, from intimidation to aggravated assault, were committed by males.

According to Cazenave and Straus (1979), youths who were violent towards their parents were more likely to come from Caucasian households than from African American or other minority households. More specifically, perpetrators were more likely to be White females than Black females, and more likely to be Anglo than Hispanic (Agnew and Huguley 1989; Paulson et al. 1990). However, considering that only a few studies have included a representative sample of racial and ethnic groups, these findings should be viewed with caution.

Studies have repeatedly documented that adolescents who assault their parents are more likely to have observed interparental violence or been abused (Brezina 1999; Browne and Hamilton 1998; Wilson 1996). Gelles (1999) revealed that one of the most consistent findings in the literature was that children who have been physically abused tend to grow up to be abusive adults. However, researchers recognize that this is probabilistic, not deterministic. In addition, prior data do not support or refute the idea that adolescents will begin expressing violence in the same environment where they learn it (Robinson et al. 1994). Although the directionality is not established, the presence of other forms of violence in the home does appear to be positively correlated to CPV (Power 1988; Straus and Gelles 1990; Wells 1987). In fact, in situations where there was wife abuse, there was also a higher pattern of abuse towards the mother by the child (Paulson et al. 1990). Severity of violence used by teenagers was also related to the severity experienced from parents and the interpersonal violence witnessed (Agnew and Huguley 1989; Browne and Hamilton 1998).

The familial environment and nature of parental relationships have been explored in the literature specific to parent assault. Families of violent adolescents were often chaotic and disorganized, and demonstrated little agreement when making decisions and choices as a group (Robinson et al. 1994; Wells 1987). Parents of violent children were often detached, distant, and disengaged from these youths (Agnew and Huguley 1989; Kratcoski 1985). Parental closeness, trust, and caring were often lacking, and parents were described as critical, unrealistic in their expectations, and lacking in adequate communication skills (Paulson et al. 1990; Wells 1987). As a result, it has been suggested that a parent’s tendency to be unrewarding, interpersonally uninvolved, emotionally detached, and permissive, may lead to inadequate social reinforcements, and subsequent antisocial practices and parent assault (Pelletier and Coutu 1992).

Individuals who commit CPV are more likely to associate with negative peer groups and approve of other forms of delinquency (Agnew and Huguley 1989; Kratcoski 1985; Paulson et al. 1990; Peek et al. 1985; Robinson et al. 1994). In fact, Evans and Warren-Sohlberg (1988) found that 66% of their adolescent sample had some history of involvement with law enforcement and social services. Thus, these friendships appeared to be reinforcing pre-existing negative beliefs about the family system, authority figures, and the like (Paulson et al. 1990).

Considering the above information, it should be of no surprise that the use of substances is highly correlated to family violence (Livingston 1986; Potter-Efron and Potter-Efron 1985). Although substance abuse is associated with family violence in general (Livingston 1986; Potter-Efron and Potter-Efron 1985), the relationship between substance abuse and CPV specifically, is indeterminate. Evans and Warren-Sohlberg (1988) reported that CPV was linked to the use of drugs or alcohol in about 20% of the cases they examined, although none reported being under the influence at the time of the assault. Straus and Gelles (1988) had different findings, noting a strong association between alcohol and family violence in their study of CPV.

School behavior has not been critically examined with respect to this population. However, researchers have reported that these youths were less likely to be interested in attending classes and more likely to describe school as unimportant (Paulson et al. 1990). In 1982, Cornell and Gelles revealed that when a child was expelled from school, violence towards a parent was more probable. Several years later, Kratcoski (1985) went further by stating that truancy, suspension, and harassment of teachers were behaviors indicative of CPV.

Little is known about the psychological characteristics of youths who are violent towards their parents. Nevertheless, studies that addressed this aspect of functioning have revealed that these children often report feeling unhappy, having low self-esteem, and possessing little self-worth (Harbin and Madden 1979; Paulson et al. 1990). According to Micucci (1995), these youths may begin to feel alienated and misunderstood and ultimately develop more significant psychological problems. Research findings clearly indicate that a child’s level of adjustment is related to the nature of their relationship with family members, particularly parents. The more dysfunctional the family relationship, the more likely these youths will develop psychological problems and engage in high-risk behaviors and the more difficult it will be for them to develop into healthy young adults.

Etiology of CPV

It has been hypothesized that the mechanisms underlying child-initiated violence within the home is distinct from other forms of juvenile crime and family violence (Agnew and Huguley 1989; Brezina 1999; Gelles 1997; Walsh and Krienert 2007). With this in mind, it is important that these youths be compared to those who have committed only nondomestic offenses in order to determine whether or not they warrant specialized treatment interventions relative to other juvenile delinquents. Child-initiated violence has been linked to factors drawn from social-learning, social-control, and social-psychological strain theories of delinquency (Agnew and Huguley 1989; Brezina 1999). Neglecting to study CPV omits a significant piece of family violence. Because so little is known about the psychological features of youths who are violent towards their parents, this study was undertaken with the primary goal of exploring their clinical and adaptive characteristics, and seeing how they differed from other juvenile offenders with no history of CPV. It was believed that a study of this nature would not only assist in determining whether or not these youths are more psychologically disturbed, but it would also provide critical information about how they can be better identified, managed, and treated.

Violent Crime

There are ongoing debates about what constitutes violence. However, for the purpose of this study, the following definition is offered: violence will refer to any behavior that involves the direct use of physical aggression towards another individual against his or her will (Emery 1989; Lystad 1986; Tate et al. 1995). Juvenile offenders in the current study classified as CPV had a history of committing at least one violent crime against a parent or caregiver (as defined by the U.S. Department of Justice’s Office of Juvenile Justice and Delinquency Prevention). Juvenile offenders who had no history (based on arrest records and interviews) of crimes against a family member were classified as NCPV.

Purpose of the Present Study

The main purpose of this study was to determine if juvenile offenders who committed violent crimes against a parent differed from those offenders who committed none. Group differences were explored with data collected on the offender’s perception of overall emotional functioning as measured by the Emotional Symptom Index (ESI) and Personal Adjustment Composite (PAC) of the Behavior Assessment System for Children (BASC)–Self Report of Personality (SRP). The profile derived from these assessments allowed for comparisons relative to overall psychological distress and adaptive functioning, as well as provided information specific to the clinical features of youths who are violent towards a parent. Additional comparisons were made relative to demographic, arrest, mental health, relationship, and intellectual findings, both within the group of CPV arrestees, and between CPV and NCPV delinquents. Youths in this study were defined as individuals aged 10 to 18.

It is important for the scientific community to further explore the phenomenon of CPV, especially when considering the (a) current CPV prevalence rates and high probability that they are significantly underestimated, (b) scarcity of studies on CPV, (c) need to further our understanding of intergenerational transmission of violence, (d) inherently disturbing implications for adjustment within the CPV relationship, and (e) limited knowledge about the underlying mechanisms of such violence.

Primary Aims

The following aims were developed to explore the clinical and adaptive features of juvenile offenders who were violent towards a parent:

• 1. General group differences. Examine variations between the CPV and NCPV group as a function of (a) demographic, (b) arrest, (c) mental health, (d) relationship, and (e) intellectual findings.

• 2. Psychological functioning. Explore the differences in the nature of emotional distress and adaptive functioning, as measured by the ESI and PAC of the SRP, between the CPV and NCPV group.

• 3. Exposure to domestic violence. Explore how youth in the CPV and NCPV group differ with regard to exposure to incidents of domestic violence.

• 4. Domestic victimization. Explore how youth in the CPV and NCPV group differ with regard to exposure to incidents of domestic victimization.

Methodology

Participants

Participants were drawn from a pool of individuals who were referred to the Juvenile Court Assessment Center (JCAC) by the Juvenile Justice Division of the Eleventh Judicial Circuit of Dade County, Florida. All participants were referred originally to the JCAC for a psychological evaluation by judges, attorneys, or other Department of Juvenile Justice personnel following an arrest.

The general sample was composed of 223 participants (168 male, 55 female; M _age = 14.55 years; SD = 1.55; age range, 10–18 years). Participants were divided by race (43.1% White, 53.8% Black, and 3.1% other races) and partitioned into five main ethnic groups: Caucasian (6.7%), Hispanic (39.9%), African American (38.1%), Haitian (9.9%), and other ethnicities (5.4%). Of these participants (N = 223), 100 had committed crimes against a parent or caregiver and 123 had not.

Instruments

The Behavior Assessment System for Children-Self Report of Personality (SRP)

The SRP was developed to evaluate the clinical and adaptive behavior of youths aged 2 1/2 to 18 years (Reynolds and Kamphaus 1998). This study used two composites of the SRP: the PAC and the ESI. The PAC is an index comprising the following factors: (a) Relations with Parents, (b) Interpersonal Relations, (c) Self-Reliance, and (d) Self-Esteem. The ESI is a composite of (a) Social Stress, (b) Anxiety, (c) Depression, (d) Sense of Inadequacy, (e) the Inverse of Interpersonal Relations, and (f) the Inverse of Self–Esteem (Flanagan 1995; Reynolds and Kamphaus 1998; Sandoval and Echandia 1994). The SRP includes three validity indexes to eliminate records that were invalid due to carelessness, inattentiveness, or cognitive limitations. They are the F index (Faking Bad), L index (Faking Good), and V index (Validity).

Coefficients of internal consistency for the BASC components range from the high .50s and low .60s to the high 80s. With respect to the SRP, internal consistencies of the scales average about .8 for males and females at both the Child and Adolescent levels. For the composites, reliabilities range from the mid-.80s to the mid-.90s, with those for the ESI ranging from .95 to .96 (Reynolds and Kamphaus 1998).

Kaufman Brief Intelligence Test (K-BIT)

The test is divided into a measure of verbal or crystallized thinking (Expressive Vocabulary and Definitions) and a measure of nonverbal or fluid thinking (Matrices). The Vocabulary subtest assesses verbal knowledge through pictures (Expressive Vocabulary) and definitions. The Matrices subtest examines ability of an individual to recognize relationships and complete analogies through pictures and abstract designs. An overall IQ is provided through the K-BIT IQ Composite. The reliabilities across age levels are more than adequate for the Vocabulary subtest (.89–.98) and the K-BIT IQ Composite (.88–.98). Although lower, reliabilities are still acceptable for the Matrices subtest (.74–.95). Test-retest reliabilities are comparable to the split-half reliabilities for the Vocabulary subtest (.86–.97), the Matrices subtest (.80–.92), and the K-BIT IQ Composite (.92–.95). Test validity was established through item analysis, and internal and external test analysis.

Wide Range Achievement Test, Third Edition (WRAT-III)

The WRAT-III was administered during the evaluation at the JAC and provided a broader estimate of achievement. The WRAT-III is a brief screening measure for achievement that covers reading recognition, spelling, and arithmetic, making it a good choice for this assessment battery. Internal consistencies are very high. Median alternate forms reliabilities are above .89, and test-retest reliabilities are at least .91. Correlations range from .50 to .70 on the California Achievement Test and Stanford Achievement Test, and .60 to .80 on the California Test of Basic Skills.

Procedures

Participants were divided into two groups: juvenile offenders who had (a) committed physical violence towards a parent (CPV), and (b) not committed violence towards a parent (NCPV). In order to validate the classification of participants, arrest histories from legal documents were examined along with interview data. Only participants whose arrest histories matched the information they gave during interviews were included in the study. Evaluations were conducted at the JCAC , or at the Juvenile Detention Center, depending on whether or not the youth was detained at the time of the assessment. Prior to conducting the evaluation, the purpose of the appointment and limits of confidentiality were discussed with the participants. Evaluations typically lasted approximately 5 h. Evaluations addressed general placement needs, treatment needs, and issues of competency to continue to trial. They involved a clinical interview, an assessment of personality, cognitive, psychoeducational, and family functioning, and a review of available academic and legal records. The child version of the SRP was completed by participants 10 and 11 years of age, while the SRP adolescence version was completed by participants 12 to 18 years of age. Resulting scores on the individual scales and composites of the SRP provided a specific measure of the dimensions of clinical and adaptive functioning. As part of the information gathering process, a psychosocial interview was conducted with a parent and a consultation was conducted with involved parties whenever possible.

Criteria for Inclusion

Participants were excluded from the study when we were unable to obtain valid arrest histories and when the SRP generated an invalid profile. Participants were also excluded from the study if they had not been in the care of a parental figure for over a year; that is, if they were residing on their own, with a significant other, or with friends. Twelve participants were removed from the general sample due to these criteria.

A parental figure refers to an adult caretaker who resides with the child and maintains the primary responsibility. This person may be a biological parent, step-parent, adoptive parent, foster parent, grandparent, other extended family members, or a combination of such. These individuals may or may not have legal custody of the child. This all-inclusive definition is necessary due to the nature of the population being studied, as research has repeatedly demonstrated that most youths within the juvenile justice system do not originate from the traditional two-parent family. It is believed that excluding these nontraditional parental figures from this study would seriously hinder the generalizability of subsequent results and conclusions. The term parent will be used interchangeably with parental figure throughout this document.

Design and Analyses

Design

A correlational approach with a cross-sectional design combined with factorial modeling was employed to determine the differences between the CPV and NCPV groups and the specifically defined subgroups of the CPV and NCPV groups. Although the ESI and PAC were the primary methods of assessing an individual’s clinical and adaptive features, information thought to be clinically relevant to the juvenile offender population was also analyzed.

Emotional distress was measured by the participant’s score on the ESI of the SRP. This is a global index of emotional distress representing the total number of negative or unpleasant emotional symptoms. Adaptive functioning was measured by the participant’s score on the PAC of the SRP.

Data Analyses

First, descriptive characteristics were measured and comparisons were made between groups (CPV and NCPV) using chi-square and t tests. These descriptive characteristics were grouped by demographic, arrest, mental health, relationship, and intellectual findings. Next, utilizing between-subjects factorial analyses, the differences in psychological functioning of the two offender types were explored. Finally, the impact of exposure to domestic violence and domestic victimization were compared by offender type.

Results

General Group Differences (Aim 1)

Demographic Findings

Gender

There were 158 males (74.9%) and 53 females (25.1%) in the total sample, with 70 males (70.0%) and 30 females (30.0%) in the CPV group and 88 males (79.3%) and 23 females (20.7%) in the NCPV group. There were no significant differences between the groups relative to gender.

Race

There were 95 Whites (45.0%), 110 Blacks (52.1%), and 6 other races (2.8%) in the total sample. When a frequency distribution was generated, it was found that there were 49 Whites (49.5%) and 50 Blacks (50.5%) within the CPV group and 46 Whites (43.4%) and 60 Blacks (56.6%) within the NCPV group. A significant difference exists between these groups relative to race for females, χ²(1, N = 52) = 4.94, p = .026, ϕ = .31 (100*[.31²] = 9.61% of the variance accounted for). Specifically, White females (76.2%) were more likely to have been violent towards a parent than Black females (45.2%).

Ethnicity

Ethnic distribution of the total sample is as follows: 14 Anglos (6.6%), 88 Hispanics (41.7%), 81 African Americans (38.4%), 17 Haitians (8.1%), and 11 Caribbean-American (5.2%). There were no significant differences between the groups.

Arrest Findings

Age of First Arrest

Mean age of first arrest for the total sample was 13.32 (SD = 1.88). There were no significant differences in the mean age of first arrest between the CPV group (M = 13.37, SD = 1.72) and the NCPV group (M = 13.27, SD = 2.02).

Arrest History

Upon reviewing available records, it was determined whether or not the individual had ever been arrested for one of the three types of offenses: violent, property, and drug. Mean number of arrests for the total sample was 3.21 (SD = 2.81), the CPV group and NCPV group were similar, receiving a mean number of 3.02 (SD = 2.56) and 3.38 (SD = 3.02) arrests, respectively. Mean number of charges for the total sample was 5.00 (SD = 5.18). For the CPV group, mean number of charges was 4.66 (SD = 4.67), and for the NCPV group, mean number of charges was 5.29 (SD = 5.60). Mean number of violent, property, and drug offenses for total sample was 1.35 (SD = 1.62), 2.25 (SD = 3.49), and .38 (SD = 1.49), respectively. It was found that youths in the CPV group (M = 1.76) had a significantly greater number of arrests for violent offenses than those in the NCPV group (M = .99), t (206) = -3.49, p < .001, d = .49. However, youths in NCPV group (M = 2.65) had more arrests for property offenses than those in the CPV group (M = 1.81), t (206) = 1.74, p = .041, d = .24.

Mental Health Findings

Psychiatric Hospitalization

When asked the question, “Have you ever been hospitalized for psychological or psychiatric reasons?,” 13.3% of the total sample responded “yes.” Results of the chi-square analysis suggest that a significantly greater percentage of participants in the CPV group (20%) reported previous psychiatric hospitalization than those in the NCPV group (7.2%), χ²(1, N = 211) = 7.48, p = .006, ϕ = .19.

Psychiatric Medication

Participants were also asked, “Have you ever been prescribed medication for psychiatric or psychological reasons?” Out of the total group, 21% responded “yes” to this question. There was a greater percentage of participants in the CPV group (29.0%) who reported being prescribed medication for psychiatric or psychological reasons than those in the NCPV group (13.6%), χ²(1, N = 210) = 7.47, p = .006, ϕ = .19.

Suicide Attempts

Suicidality was assessed during the course of data collection, and it was found that 10.9% of the total sample had attempted suicide. Youths in the CPV group (19.0%) were significantly more likely to report a previous suicide attempt than those in the NCPV group (3.6%), χ²(1, N = 211) = 12.84, p < .001, ϕ = .25.

Relationship Findings

Poor Family Relationships

Participants were asked to respond “yes” or “no” to whether or not they found their relationships with other household members to be poor. The aim of this question was to provide a quick and simplified measure of how the youth perceive their family relationships. Additionally, the BASC Relations With Parents scale was administered to provide a more systematic and detailed description of the youth’s relationship with his or her parent. In the total sample, 29.5% of the participants described their relationships with household members as poor. Youths in the CPV group (44.4%) were significantly more likely to report poor relationships with family members than youths in the NCPV group (16.25%), χ²(1, N = 210) = 20.04, p < .001, ϕ = .31. On the BASC “Relations With Parents” scale, the mean score for the total sample was 43.49 (SD = 13.29). Results revealed that youths in the CPV group (M = 40.63) were significantly more likely to have reported difficulties with parents than those in the NCPV group (M = 46.07), t (209) = 3.03, p = .003, d = .42.

Negative Peer-Group Involvement

Participants were asked whether most, some, or none of their peers commit crimes, own guns, use drugs, or are gang members. Youths were also asked to respond “yes” or “no” to whether or not they are gang members or affiliate with gang members. The data suggest that most youths in the (a) total sample (62.4%), (b) CPV group (64.9%), and (c) NCPV group (60.3%) had peers who commit crimes. In addition, youth in the CPV group were significantly more likely than those in the NCPV group to have reported that their peers own guns, χ²(1, N = 208) = 6.85, p = .033, ϕ = .18. Additionally, they were more likely to have reported that their peers belong to a gang, χ²(1, N = 210) = 6.29, p = .043, ϕ = .17. Furthermore, youths in the CPV group were significantly more likely than the NCPV group to have reported that they affiliate with gang members, χ²(1, N = 211) = 3.93, p = .048, ϕ = .14.

Parental Figure

Table 1 was generated to provide the reader with information about the total sample with respect to this variable and as a way of directly comparing the frequency distributions and percentages of parental figures for the CPV and NCPV groups. In many respects, the groups were quite similar, with the highest percentage of youths from the CPV group (39.0%), NCPV group (33.3%), and total sample (36.0%) residing in a single parent home with their biological mother as the primary caretaker. However the CPV group was about half as likely to live with both biological parents compared to the NCPV group.

Table 1 Frequency and percentage distribution of parental figures of the CPV and NCPV groups

Intactness of Household

Participants were asked whether or not both parents reside in the home. If a youth responded no, then he or she was asked to provide the clinician with the physical location of each parent. For descriptive purposes, these categories were collapsed to represent intact versus not intact. Intact is defined by the categories “married, living together” and “never married, living together.” The remaining categories encompass the not intact group. With respect to the total sample, 15.6% reported that they came from an intact family and 84.4% reported that they came from a household that was not intact. Although there was no statistically significant difference between the groups with respect to “intactness of household,” χ²(1, N = 211) = 3.10, p = .078, ϕ = .12, youths in the CPV group (11.0%) were less likely to have come from an intact home than youths in the NCPV group (19.8%).

Intellectual Findings

IQ

The K-BIT was used to assess intelligence. The mean score on the K-BIT Composite for the total sample was 78.20 (SD = 16.00). The mean score for the CPV group was 79.35 (SD = 16.01) and for the NCPV group, 77.17 (SD = 15.99). The t-test results suggest that no significant differences exist between participants in the CPV and NCPV groups relative to IQ.

Achievement

Achievement was assessed by the participants (a) scores on the Reading subtest of the Wechsler Reading and Achievement Test–Third Edition (WRAT-3), (b) record of any grade failure, and (c) placement in special classes. Information for the latter two variables was provided by the youth and then verified by parental figures and school records. The mean scores and standard deviations for the total sample (M = 83.17, SD = 19.78), CPV group (M = 83.49, SD = 20.75) and NCPV group (M = 82.87, SD = 18.75) on the WRAT-3 Reading subtest were quite similar, as were their mean grade levels, (M = 6.39, SD = 4.35), (M = 6.56, SD = 4.50), and (M = 6.25, SD = 4.22), respectively.

School Performance

The data show that 51.7% of the participants in the total sample reported having failed a grade. This is similar to the percentage of participants in the CPV (51.0%) and NCPV (52.3%) groups. When asked if they were in special classes, 33.5% of the total sample responded “yes.” In the CPV group, 27.3% responded “yes,” while in the NCPV group, a higher percentage (39.1%) said “yes.” Participants within the NCPV group (39.1%) were more likely to have been placed in special classes than the CPV group (27.3%), χ²(1, N = 209) = 3.27, p = .05, ϕ = .13.

Psychological Functioning (Aim 2)

As Table 2 reveals, mean scores on the ESI for participants in the CPV and NCPV group are quite similar, as are the mean scores for males and females within these groups. No significant differences in the ESI scores were found by group, F(1, 207) = .57, p = .45, η² = .003 , by gender, F(1, 207) = .178, p = .184, η² = .008, or as an interaction of group by gender, F(1, 207) = .12, p = .782, η² = .001 (see Table 3).

Table 2 Means and standard deviations of scores on the emotional symptom index for the total sample, CPV group, and NCPV group, by group and gender

Table 3 Two-way analysis of variance for scores on the emotional symptom index for the CPV and NCPV groups by gender, race, and ethnicity

The PAC assesses a youth’s positive level of adjustment with respect to relations with parents, interpersonal relations, self-reliance, and self-esteem. As Table 4 reveals, mean scores and standard deviations on the PAC for participants in the CPV group and NCPV group, by group and gender appear to be quite similar to one another, with females in the NCPV group receiving the highest scores (M = 49.17, SD = 10.34) and females in the CPV group receiving the lowest scores (M = 45.03, SD = 11.82). There were no significant main effects for group, F(1, 207) = .2.29, p = .132, η² = .01 or gender, F(1, 207) = .54, p = .463, η² = .003, or any significant interaction effect of group by gender, F(1, 207) = 1.36, p = .245, η² = .006 on the PAC (Table 5).

Table 4 Means and standard deviations of scores on the personal adjustment composite for the total sample, CPV group, and NCPV group, by group and gender

Table 5 Two-way analysis of variance for scores on the personal adjustment composite for the CPV and NCPV groups by gender, race, and ethnicity

Race

The mean scores and standard deviations on the ESI for participants in the total sample, CPV group, and NCPV group by race appear similar to one another. Results of the ANOVA are presented in Table 3 and reveal that there were no significant main effects for group, F (1, 201) = .52, p = .472, η² = .003 or race, F (1, 201) = 2.00, p = .159, η² = .01; however, there was a significant interaction effect for group by race, F (1, 201) = 2.94, p = .044, η² = .02. For the CPV group, Blacks received a higher ESI score (M = 53.12, SD = 10.72) than Whites (M = 48.90, SD = 9.01).

Mean scores and standard deviations on the PAC by participants in the CPV group and NCPV group, by group and race, reveal that Whites in the NCPV group received the highest scores (M = 49.78, SD = 8.86) and Blacks in the CPV group received the lowest scores (M = 46.04, SD = 11.01). Table 5 presents the results and suggests that there were no significant main effects for group F(1, 201) = 1.52, p = .219, η² = .007 or race F(1, 201) = 2.37, p = .125, η² = .01 or any significant interaction effect for group by race F(1, 201) = .03, p = .866, η² = .000.

Ethnicity

Mean scores and standard deviations on the ESI for participants in the total sample, CPV group, and NCPV group, by group and ethnicity, reveal that Haitians in the CPV group received the highest scores (M = 54.17, SD = 11.65), while Haitians in the NCPV group received the lowest scores (M = 47.73, SD = 8.99). Results presented in Table 3 reveal no significant main effects for group F(1, 201) = .37, p = .545, η² = .002 or ethnicity F(1, 201) = .48, p = .751, η² = .01 or significant interaction for group by ethnicity F(1, 201) = .87, p = .481, η² = .02.

Mean scores and standard deviations on the PAC for participants in the total sample, CPV group and NCPV group, by group and ethnicity, revealed that Anglos in the NCPV group received the highest scores (M = 52.50, SD = 6.16) and Caribbean Americans in the CPV group received the lowest scores (M = 39.80, SD = 16.65). The results presented in Table 5 reveal no significant main effects for group F(1, 201) = 2.20, p = .14, η² = .01 and ethnicity F(1, 201) = .85, p = .496, η² = .02, and no significant interaction effect for group by ethnicity F(1, 201) = .48, p = .753, η² = .009.

Exposure to Domestic Violence (Aim 3)

In order to assess exposure to domestic violence, youths were asked to respond “yes” or “no” to the question, “Have you ever witnessed any form of violence, verbal or physical, between members of the household?” Most youths in the total sample (66.8%) and NCPV group (82.9%) had never witnessed family violence. However, youths in the CPV group were more likely to report witnessing family violence (51.0%) than to report that they had not witnessed such violence (49.0%). The chi squared analysis revealed that youths in the CPV group (51.0%) were significantly more likely to have been exposed to incidents of domestic violence than those in the NCPV group (17.1%), χ² (1, N = 211) = 27.24, p < .001, ϕ = .36.

Domestic Victimization (Aim 4)

In order to assess “domestic victimization,” youths were asked whether or not they had ever been physically or verbally victimized by another household member. Most youths in the total sample (60.2%) and NCPV group (81.1%) reported no victimization by a family member; however, most of the youths in the CPV group (63%) reported victimization. The results of the chi square analysis revealed that youths in the CPV group (63%) were significantly more likely to report domestic victimization than those in the NCPV group (18.9%), χ² (1, N = 211) = 42.66, p < .001, ϕ = .45.

Discussion

Several of the analyses conducted in this study yielded significant results, suggesting that young offenders who are violent towards a parent differ from the general population of young offenders. The few published studies that examined the psychological and psychiatric functioning of youths who committed acts of parent violence, found that these youths typically felt alienated, misunderstood and unhappy, reported having low self-esteem and little self-worth, and presented with higher levels of family dysfunction and poorer family relationships (Cornell 1990; Paulson et al. 1990). This study revealed significant results that were very interesting with respect to psychological functioning. It was found that youths in the CPV group were more likely than children in the NCPV group to have a history of: (a) psychiatric hospitalization, (b) psychotropic medication use, and (c) attempted suicide. These results strongly suggest that juvenile offenders who are violent toward a parent are more psychologically disturbed than those offenders who are not.

The literature regarding the relationship between gender and age and parental violence were conflicting (Walsh and Krienert 2009). While some researchers found that males were more likely to be violent towards a parent than females, others found no relationship between the sex of the child and the probability of violence towards a parent (Browne and Hamilton 1998; Paulson et al. 1990). In the present study, males and females were equally as likely to have a history of violence towards a parent. As for race, our findings were in agreement with the literature indicating that youths who were violent towards their parent were more likely to come from Caucasian households than from other minority households (Agnew and Huguley 1989; Paulson et al. 1990). Specifically, White females were more likely than Black females to have been arrested for violence towards a parent. Additionally, we found that youths in the CPV group were less likely to have come from an intact home than those in the NCPV group, suggesting that domestic violence against a parent may be related to intactness of family.

Interestingly, when arrested, most of the youths in the CPV and NCPV groups were residing in a single-parent home with only the biological mother as the caretaker, suggesting that youths who have contact with the juvenile justice system typically come from a single-parent home. This is consistent with the literature that indicates youths from single-parent homes have higher rates of problem behaviors, aggression, and delinquency than those who come from two-parent homes (Griffin et al. 2000; Hoffman 1993; Van-Kiernan et al. 1995).

The nature of parental relationships explored in studies specific to CPV have indicated that parental closeness, trust, attachment, and caring were often lacking, and that parents were often perceived as critical, unrealistic in their expectations, and lacking in adequate communication skills (Paulson et al. 1990; Peek et al. 1985; Wells 1987). This study supported these findings, as youths in the CPV group were more likely to report difficulties relating to their parents as assessed by the BASC-Relations with Parents scale than those in the NCPV group.

Results of this study yielded no significant differences between groups on total number of arrests or charges incurred by the offender. However, youths in the CPV group were significantly more likely to have been arrested for other violent offenses than those in the NCPV group, while youths in the NCPV group were more likely to have been arrested for property offenses than those in the CPV group. Both groups were equally as likely to have been arrested for a drug offense. These findings are important, as they highlight the suggestion that children who are violent towards their parent may be more violent outside of the home than those who are not violent towards a parent. Additionally, it appears that the CPV group is equally as likely to engage in other illegal behaviors as their nondomestic offending counterparts. Furthermore, these findings indicate that CPV may not only be the product of dysfunctional family relationships, but also the upshot of other behavioral expressions of a conduct-disordered child.

Adolescents who were violent towards their parents have often observed interparental violence (Brezina 1999; Browne and Hamilton 1998; Gelles 1999; Paulson et al. 1990; Straus and Gelles 1990; Wilson 1996). In addition, a high percentage of violence towards parents is bidirectional and the severity of violence appears to be related to the severity experienced from parents and the interpersonal violence witnessed (Agnew and Huguley 1989; Browne and Hamilton 1998). Given these findings, it was hypothesized that youths who were violent towards a parent would be more likely to report having been exposed to incidents of domestic violence than those who were not. This hypothesis was supported, as youths in the CPV group were significantly more likely to report domestic violence exposure than those in the NCPV group

Studies have also consistently documented that youths who assault their parents are more likely to have been abused themselves (Brezina 1999; Browne and Hamilton 1998; Gelles 1999; Wilson 1996). The literature established that parents of conduct-disordered or difficult-to-manage children tend to be more aggressive when dealing with their offspring (Bradley and Peters 1991; Conger et al. 1995; Vuchinich et al. 1992). As expected, a greater percentage of youths in the CPV group reported a history of abuse when compared to youths in the NCPV group.

Summary

In many ways, juvenile offenders who were violent toward a parent were quite distinct from those who were not. Familial variables examined in this study clearly differentiated youths who assault their parents. It was found that these youths were less likely to have come from an intact home and that they were more likely to have difficulties relating to their parents and to other household members. They were also more likely to have been exposed to domestic violence and victimized by family members. When their peer groups were examined, the youths in this study were found to be quite similar with respect to having friends who commit crimes and use drugs. However, youths who were violent towards a parent were more likely to have peers who own guns and belong to a gang. Also, they appeared to be more likely to affiliate with gang members overall. The arrest histories revealed that both groups incurred their first and current arrest at approximately the same age. However, youths who were violent towards a parent committed a greater number of violent offenses, while those who were not violent towards their parent committed a greater number of property offenses. Youths who were violent towards a parent seemed to experience greater psychological distress than those who were not. They were more likely to have been psychiatrically hospitalized and medicated, and more likely to have attempted suicide. Females reported more psychological distress than males overall, and African American youths who were violent towards their parent reported greater psychological distress than Whites who committed such violence. Finally, a higher percentage of the youths in the CPV group were placed in classes to address emotional and behavioral problems, while a higher percentage of youths in the NCPV group were placed in classes to address learning problems.

Limitations of the Present Study

First, although this study contained a reasonable number of participants who had been arrested for violence towards a parent (n = 100), a larger sample would have been desirable because it would have allowed for larger cell sizes during analyses and greater overall power. The second limitation has to do with selection bias. Recall that the only youths included in this study were those who had been referred to the JCAC for an evaluation by an attorney or judge. Thus, it was the legal systems subjective decision alone that determined whether or not the youth would be excluded from participating. Third, although the very nature of a court-ordered evaluation may have had a positive impact of data verification, there were some difficulties within the court setting that may have negatively affected data collection. The political climate and poor interdepartmental communication made it difficult at times to obtain pertinent information. Moreover, the information reported by caretakers during the course of the evaluation may have been biased in light of impending legal sanctions.

Future Directions

More in-depth studies are necessary before the professional community can begin to make clear-cut assessment and treatment recommendations for delinquent youths who assault a parent. First and foremost, researchers should address the lack of agreement with respect to fundamental operational definitions used in the CPV literature. Without more conformity, it is difficult for researchers and clinicians alike to clearly understand the key aspects of the field, compare and contrast studies, generalize results, and effectively utilize findings. Researchers should also begin to address the psychological features (e.g., examine the differential impact of various predictor variables) of these youths more comprehensively so that the professional community is provided with theoretically and clinically useful information to identify, treat, and dispose of such cases. Without this information, it is difficult for the professional community to appropriately manage victims and perpetrators when they come to the attention of social agencies and law enforcement.