Adolescents at risk for violence: an initial validation of the life challenges questionnaire and risk assessment index.
Adolescent violence is a serious issue which has gained attention nationally. Such problems as gang violence, teen-on-teen homicide, aggravated sexual assault, and domestic battery involving children and adolescents have raised concern among policymakers, law enforcement officials, school administrators, and others. Despite the nation's increased interest and concern, youth violence is not a new phenomenon. According to the 1999 "National Report on Juvenile Offenders and Victims" produced by the Office of Juvenile Justice and Delinquency Prevention (OJJDP) (1999), the proportion of violent and seriously violent crimes (rape, sexual assault, robbery, aggravated assault, and simple assault) committed by juveniles has remained relatively constant since the 1980s. However, the total annual number of homicides committed by juveniles doubled between 1987 and 1994, with the majority committed by African American males between the ages of 15 and 17. Though the rate of juvenile homicide has since decreased, it remains at a level that is 21% above that of the 1980s (OJJDP, 1999). The Federal Bureau of Investigation's Supplemental Homicide Report, based on reported homicides nationwide, indicates that 12% of all murders committed nationwide involve at least one juvenile offender (OJJDP, 1999). Furthermore, recent research has estimated that the average career criminal, heavy drug abuser, and high school dropout costs society $3 million dollars in medical, legal, and psychological services, lost wages, and justice system costs (Cohen, M. as cited in OJJDP, 1999). It is clear from these statistics that juvenile violence and delinquency remains a major problem. In addition, the United States has the highest overall rate of violence among industrialized nations, with a significant surge of violent and aggressive acts being committed during the teen years (Dusenbury, Falco, & Lake, 1997).
As a result, a great deal of research in the past decade has focused on identifying trends in youth violence and risk behaviors in order to gain a better understanding of the magnitude of the problem and develop strategies for intervention and prevention. For example, in 1990 the Centers for Disease Control and Prevention (CDC) developed the Youth Risk Behavior Survey (YRBS), a self-report inventory designed to track risk behaviors of youth in grades 7 through 12 (Dorman & Pealer, 1999). This self-report questionnaire consists of 75 forced-choice items administered in the regular classroom using standard Scantron answer sheets. Since its inception, the YRBS has been administered every two years to a nationwide random sample of middle school and high school students in an effort to monitor the incidence and prevalence of six categories of risk behaviors including injuries, tobacco use, alcohol and other drug use, sexual behavior, dietary behaviors, and physical activity (Dorman & Pealer, 1999). In a recent report based on this survey, 20% of students nationwide brought a weapon to school, 38.7% had been in physical fights, 53.1% engaged in sex, 32% reported heavy drinking, 71% tried tobacco, 42.4% reported marijuana use, and 7% indicated use of cocaine in some form (Stevens & Griffin, 2001). In 1997, the National Crime Victimization Survey, a self-report questionnaire administered to a random sample of adolescents above the age of 12, indicated that adolescents were involved in 25% of serious violent victimizations annually, including rape, sexual assault, robbery, and aggravated assault (OJJDP, 1999). Despite increasing interest in assessing the magnitude of adolescent involvement in violence-related crime and delinquency and monitoring changes in these activities, less attention has been focused on prevention efforts through early identification of adolescents in the general population who are at risk for involvement in violence. Early identification of adolescents at risk for violence is needed so that clinicians, educators, parents, and others can channel resources to assist adolescents in coping with challenges in their lives before they become involved in serious offenses.
The purpose of the present study is to examine the utility of a screening instrument and assessment scale (The Life Challenges Questionnaire-Teen Form--LCQ-TF and The Risk Assessment Index-RAI) in identifying young people at risk for violence by administering the instrument to a sample of adolescents in a juvenile detention population and a sample of adolescents attending a private high school. The specific objective is to compare the two groups of adolescents--students in a high school and a detention setting--in order to identify challenges faced during adolescence and to provide initial validation data for a scale to specifically assess risk for violence among adolescents.
The Life Challenges Questionnaire-Teen Form (LCQ-TF) is a 120-item, self-report inventory developed by licensed clinical psychologists at the Center for Applied Psychology and Forensic Studies, a human services corporation in Chicago, Illinois. The LCQ-TF was designed with the specific purpose of providing a practical toot to help clinicians, educators, and parents identify children and adolescents who are coping poorly with the challenges of their lives and are at risk for involvement in violent behavior. The questionnaire was developed based on previous research establishing risk factors that strongly correlate with youth violence and aggression. Each item asks respondents about important aspects of their life, including family structure, parental relationships, sibling and peer relationships, neighborhood environmental context, academic performance, school connectedness, life conflicts, gang memberships, substance abuse, physical and sexual abuse, trauma, and other factors. While all items reflect known correlates of youth violence, there are 53 critical items scattered throughout the questionnaire. The Risk Assessment Index (RAI) is based on these 53 critical items found to be most strongly correlated with youth violence. These items relate to the most strongly correlated predictors of youth violence based on previous research. These items include questions about gang membership, attachment to and perceptions of caregivers, relationships among family members, the presence of violence or discord in the home, personal attitudes about aggression and violence, school connectedness and involvement, quality of the school environment, substance abuse in the home and among peers, self-esteem and self-perceptions, peer involvement and connectedness, involvement in recreational activities, recent loss or trauma, and any history of physical or sexual abuse. Additionally, these items assess for any psychological or emotional difficulties, suicidal or homicidal thoughts, and engagement in antisocial behaviors such as fighting, assault, harming animals, and the use of or access to weapons.
REVIEW OF THE LITERATURE
Research on Risk Factors for Adolescent Violence
Over the past decade a great deal of research has been conducted in an effort to understand the causes and correlates of juvenile violence as well as to help develop effective early identification and prevention programs. Much of this research has focused on the identification of predisposing risk factors for violence and the development of reliable and valid violence risk assessment methodologies. Various epidemiological studies based on data compiled from government agencies by the OJJDP and other researchers have shown that a greater percentage of seriously violent crimes are committed by boys, that violence peaks between the ages of 15 and 17, that juvenile violence peaks in the after-school hours between 3 P.M. and 4 P.M. on school days and in the evenings between 8 P.M. and 10 P.M on non-school days, and that prevalence rates of violent behavior are higher in minority populations (OJJDP, 1999; Kelley, Huzinga, Thornberry, & Loeber, 1997; Rachuba, Stanton, & Howard, 1995). This same research has also shown that violent behavior tends to occur intermittently rather than continuously, with 75% of violent offenders reporting sporadic offending over a 5-year period (Kelley, Huzinga, Thornberry, & Loeber, 1997).
Recent research has also identified several predisposing characteristics that are highly correlated with later violence among children and adolescents. These risk factors tend to cluster around four main areas of the juvenile's life--the family system, the school, personality and psychological factors, and the peer group. Risk factors are relatively consistent across studies.
Research conducted by Bischof, Stith, and Whitney (1995) sought to compare the family environments of adolescent sex offenders and violent and non-violent juvenile delinquents with those of a normative sample of adolescents. Participants in this study included 105 adolescent males between the ages of 12 and 18 in outpatient and residential programs. Of the 105 participants, 39 were classified as sex offenders based on self-reported child sexual molestation or involvement in a sex offenders treatment program, 25 were classified as violent offenders based on self-reported engagement in violent offenses (homicide, manslaughter, robbery, and aggravated assault), and 41 were classified as non-sex offenders based on self-reported engagement in non-violent offenses (those against persons or property, status offenses, and substance abuse violations). Researchers administered the Family Environment Scale Form-R (FES), a true-false self-report instrument measuring social and environmental characteristics of various types of families. Specifically, this instrument measures the adolescent's perceptions of family cohesion, emotional expressiveness, and conflict as well as the degree of emphasis on independence, political, social, intellectual, recreational, and cultural activities. The FES also assesses the degree of emphasis on religion, ethics, and morality as well as the level of family organization (roles, structure, responsibilities) and dependence on rules and procedures. The results of this study suggest that family environments characterized by emotional disengagement, a lack of perceived cohesion among family members, and a strong reliance on parental control and rules correlate with adolescent violence (Bischof, Stith, & Whitney, 1995). Moreover, delinquent, behavior is also correlated with family environments characterized by a lack of interest in political, social, intellectual, cultural, and recreational activities. Finally, children and adolescents who lack reasonable, age-appropriate independence and privacy within their family environments tend to have higher rates of violence.
A later study, conducted by Rodney, Tachia, and Rodney (1999), supports many of these previous findings and also sheds light on several other important family characteristics associated with delinquency. In this study, researchers sought to identify the relationship between delinquency and the family environment, including such factors as the relationship with extended family, the amount of time parents spent with their children, the manner of discipline used, the nature of family rules, the relationship with siblings, and the adolescent's peer network. The researchers interviewed 556 African American adolescents between the ages of 13 and 17 living in eight different states across the nation. The researchers sampled participants from neighborhoods known to have high rates of drug use and violence. All were enrolled in violence, alcohol, and drug prevention programs. Participants were each administered sections of the Children's Structured Assessment for the Genetics of Alcoholism (C-SAGA) pertaining to demographic information, the child's environment, and behavior problems associated with conduct disorder (Rodney, Tachia, & Rodney, 1999). Results indicated that all of the aforementioned familial characteristics were correlated with delinquency, particularly lack of a stable father-figure, lack of involvement of extended family, lack of adult extrafamilial supports, harsh disciplinary practices coupled with inconsistent rules and confused roles, and negative peer influences (Rodney, Tachia, & Rodney, 1999).
Several studies have also focused on extra-familial risk factors for youth violence. For example, Bonny, Britto, Klostermann, Hornung, and Slap (2000), administered a modified version of the National Longitudinal Study of Adolescent Health to 3,497 urban and suburban public school students in grades 7 through 12. This study sought to identify the relationship between school connectedness, defined as the student's perception of his or her closeness, belongingness, happiness, fairness, and safety at school, and high-risk behaviors. Results do suggest a significant relationship between school disconnectedness and high-risk behaviors. Additionally, Flannery, Singer, and Wester (2001), sought to compare levels of violence exposure, violent behaviors, psychological trauma, and suicide risk between a sample of dangerously violent adolescents and a matched sample of non-violent adolescents. Participants included 3,735 high school students between the ages of 15 and 18 at six suburban, urban, and small city public high schools in Ohio and Denver. Dangerously violent (DV) youth were defined as those who reported attacking or stabbing someone with a knife or shooting at someone with a real gun in the past year on a self-report survey. Comparison students, who did not report engagement in any type of violent behavior, were matched to the DV sample based on age, gender, ethnicity, area of residence, and family structure. Exposure to violence, engagement in violent behavior, and psychological trauma were assessed using self-report inventories. Results suggest a significant relationship between exposure to violence, violence victimization, psychological trauma, and later dangerously violent behavior.
Finally, a more recent study conducted by Ellickson and McGuigan (2000) has perhaps provided the most comprehensive, clinically sound, and robust data on adolescent risk factors. In this study, the researchers sought to identify early predictors of adolescent violence and to assess whether these risk factors varied by sex and different types -and levels of violence. These researchers gathered data from a 5-year longitudinal self-report survey of more than 4,300 high school seniors and dropouts in the states of California and Oregon. Subjects in the study were participants in the RAND Adolescent Panel Study, a longitudinal survey of substance use and health-compromising behaviors. Students were originally drawn from 30 California and Oregon junior high schools for participation in the RAND study. The sample is thought to be representative of adolescents in general, as the schools were within urban, suburban, and rural districts. Moreover, 9 of the schools had 50% or more minority populations, and 18 schools were within low-income areas. The current study used data from the first and seventh data collection waves between 1985 and 1990. Students were first administered the RAND survey in 7th grade to determine baseline statistics on the aforementioned behaviors. These same students were contacted by mail 5 years later to fill out a second survey assessing engagement in various forms of violence. At the time of the second survey, all students would have been in the 12th grade. Of the 4,390 students who completed this additional survey, 10.4% had dropped out of school, 71% were White, 8% were African American, 9% were Hispanic, 9% were Asian, and 3% were Native American or students identified as having other racial/ethnic backgrounds.
The researchers used the data from this second survey to regress measures of relational, predatory, and overall violence on predictors measured 5 years earlier (Ellickson & McGuigan, 2000). The dependent variables consisted of 3 types of violence assessed at the 12th grade, including relational violence (hitting or threatening a family member or someone outside the family), predatory violence (past-year use of force or strong-arm methods to obtain money or things from people, involvement in gang fights, attacking someone with the intent of seriously harming or killing them, and/or carrying a hidden weapon), and overall violence (includes all 6 of the violence types). The study examined predictors of violence for each category and the amount of each type of violence in the previous year. Predictor variables covered 7 domains including poor school bonds, disrupted and disengaged family bonds, other problem behaviors like early deviance and substance abuse, exposure to deviant behaviors, exposure to deviant social influences including prevalence of alcohol, tobacco, and drug abuse among friends and peers as well as the number of offers the student received for the substance, personality and attitudes including low self-esteem and rebelliousness, school and neighborhood context including prevalence of drug use among 8th graders at the student's middle school and socioeconomic status of the school's catchment area, and sociodemographic characteristics including the students age at baseline, sex, parental education, and race/ethnicity.
Results of the study suggested several very interesting patterns among adolescent risk behaviors that tend to predict future violence. All of the predictors were significantly related to the occurrence of overall violence except racial/ethnic identification as Hispanic, Asian, Native America, or Multiracial. However, being Hispanic was a significant predictor of later predatory violence. Age was not a significant predictor of relational violence. The presence of poor academic performance, early deviant behavior, and being male in the 7th grade consistently predicted all types of future violence in 12th grade. Attending a middle school with a relatively high level of drug use and moving from one school to another significantly increased the occurrence of overall and relational violence in the 12th grade. Low self-esteem predicted both relational and predatory violence. Rebelliousness did not predict any type of later violence. Being older than most students for one's grade level served as a protective factor, decreasing the risk of later relational violence.
Drug use in the adolescent's middle school was the only variable that predicted the amount of relational violence 5 years later while poor grades, high elementary school mobility, early deviance, and sex predicted the amount of overall violence. A high frequency of using alcohol, tobacco, and drugs in grade 7, higher levels of drug use among one's friends, being male, being multiracial, living in a traditional nuclear family household with two parents, and a lack of rebelliousness all predicted later predatory violence. The study also suggested that both relational and predatory violence are fostered by early deviance, doing poorly in middle school, and low self-esteem, while being female serves an inhibitory function. Unique predictors of relational violence included attending more than two elementary schools and attending middle schools with a higher prevalence of drug use among students. Unique predictors for predatory violence included early drug use and high levels of drug use among one's friends during middle school.
Several predictors of violence also varied significantly by sex. Engaging in early deviant behavior and attending middle schools with a high level of drug use had similar effects for boys and girls. However, having poor grades in middle school predicted increased relational violence for girls and increased predatory violence for boys. For girls, having low self-esteem predicted later relational violence and living in low socioeconomic neighborhoods predicted predatory violence. For boys, attending more than one elementary school and being offered drugs during 7th grade predicted future relational violence. Being Caucasian decreased later predatory violence for boys and girls but decreased later relational violence among girls only. Overall, the frequency and range of violence was lower for girls than boys.
Taken together, these studies suggest several risk factors that are highly predictive of adolescent violence. Family characteristics associated with youth violence include rigidity or an inability to adapt well to changes, enmeshment among parents and children where there is little privacy or independence for the adolescent, and a great deal of chaos and role confusion (Rodney, Tachia, & Rodney, 1999; Ellickson & McGuigan, 2000). Furthermore, emotional disengagement among family members, a lack of perceived cohesion and rigid parental control and rules also correlate with youth violence (Bischof, Stith, & Whitney, 1995). School-related correlates of youth violence include poor academic performance, weak ties or bonds with schools, high school mobility, and poor school environments (Elickson & McGuigan, 2000; Bonny, Britto, Klostermann, Hornung, & Slap, 2000). In particular, children and adolescents showing little interest in school and minimal involvement in school-related activities during the middle school years are more likely to engage in later violent behavior (Bonny et al., 2000). Poor school environments include high levels of alcohol, tobacco, and drug abuse in the student population, high rates of gang activity, and a perceived lack of closeness to school personnel. Middle schools with a higher percentage of alcohol, tobacco, and drug abuse have higher rates of violence among their students (Elickson & McGuigan, 2000). Similarly, early drug use and involvement with a drug-abusing peer group also increases the risk of later violent behavior. Such mental health factors as depression, low self-esteem, trauma, and loss also correlate with violent and aggressive behaviors (Ellickson & McGuigan, 2000; United States Secret Service, 2000; Flannery, Singer, & Wester, 2001).
Preventing Adolescent Violence: The Need for Early Identification of Risk Behavior
Although this research has greatly advanced our knowledge and understanding of youth violence, the problem continues with no clear-cut, professionally agreed upon solution. However, one factor remains consistent throughout' the literature. Study after study has shown the importance of early identification of high-risk youth in preventing future violent behavior. Consequently, the United States Department of Education (1999) produced a manual to assist school administrators, teachers, families, and students in creating a safe school environment. This manual is based upon a compilation of data and information gathered from several national organizations on school safety as well as clinical studies such as those mentioned above. Throughout the manual, the authors suggest that early identification, through evaluation and assessment of risk factors and warning signs, is the only effective means of violence prevention, as it allows children to get necessary treatment before it is too late. The United States Secret Service's Safe School Initiative (2000) also highlights the fact that there are almost always early warning signs that are ignored or dismissed prior to youth violence. Researchers and other professionals have also suggested that the success of any violence prevention program depends upon early identification of high-risk youth in order to ensure that they receive appropriate services. For example, Guetzloe (2001) suggests that many of the risk factors related to violence can be successfully addressed. However, education, skills building, co-operation between families, schools, and professionals, staff training, and, most importantly, early identification and intervention with children with emotional and behavioral problems are essential for successful violence prevention. Astor (1995) discusses the stability of early aggressive behavior across the lifespan, emphasizing the importance of identifying seriously aggressive children and intervening very early on, perhaps during kindergarten or 1st grade. The literature clearly and consistently supports early identification.
However, previous efforts to develop and validate clinically sound actuarial risk assessment tools have failed to provide reliable, dependable, and effective measures of violence risk. Research efforts have often been flawed by serious methodological problems including, complex and difficult assessment procedures, the consideration of a limited range of risk factors, weak measures of violence such as arrest histories, and samples with limited demographic, diagnostic, personality, and historical variability (MacArthur Research Network on Mental Health and the Law, 2001). For example, one study seeking to establish criterion validity of the Social Symptomatology Scale, a measure of antisocial behavior, on the Holden Psychological Screening Inventory was conducted on a limited sample of 214 adult prison inmates in Canadian Federal Correctional Facilities (Book, Knap, & Holden, 2001). While the results lend support to the efficacy of this measure, the sample size is too restricted to provide conclusive or definitive validation. Another study assessing the structure of the Self-Restraint Scale, a predictive measure of problem behaviors among adolescents, on the Weinbeger Adjustment Inventory, sampled African American 6th graders from mostly low income, single-parent families (Farrel & Sullivan, 2000). Results of this study were also encouraging, suggesting that the structure of this scale provides a relatively accurate assessment of socioemotional adjustment. However, the demographic variables are too limited to generalize the results to the general population. The MacArthur Violence Risk Assessment Study (2001), considered one of the most extensive and clinically sound risk-assessment studies to date, assessed the efficacy of a comprehensive tree-based assessment model. This study sampled a diverse range of individuals, with various diagnostic, socioeconomic, personality, and demographic variables. However, participants were all patients hospitalized in inpatient acute-care psychiatric facilities. The researchers asssessed multiple outcome variables and risk factors. Results suggest that this model provides an effective and accurate means of violence risk assessment within a very restricted population. Moreover, the methodology is extremely complex and time-consuming, limiting its practicality for use within more general clinical practice, schools, or the juvenile justice system. Thus, clinicians, parents, and educators remain burdened by the lack of a practical, reliable, and valid risk-assessment methodology. The present study seeks a solution to these difficulties.
In summary, the literature indicates that, despite an increase in serious violence-related offenses committed by juveniles, there continues to be a lack of attention paid to developing effective assessment tools for early identification of adolescents at risk for violence. The present study addresses this shortcoming by testing the utility of the Risk Assessment Index, which was developed for this purpose. As previously mentioned, the Risk Assessment Index (RAI) is a scale based on 53 critical items found in the Life Challenges Questionnaire--Teen Form (LCQ-TF). These items relate to the most strongly correlated predictors of youth violence based on previous research. These items include questions about gang membership, attachment to and perceptions of caregivers, relationships among family members, the presence of violence or discord in the home, personal attitudes about aggression and violence, school connectedness and involvement, quality of the school environment, substance abuse in the home and among peers, self-esteem and self-perceptions, peer involvement and connectedness, involvement in recreational activities, recent loss or trauma, and any history of physical or sexual abuse. Additionally, these items assess for any psychological or emotional difficulties, suicidal or homicidal thoughts, and engagement in antisocial behaviors such as fighting, assault, harming animals, and the use of or access to weapons. In the present study, the Risk Assessment Index is administered to adolescents in detention and non-detention populations to examine its potential utility as a screening tool in assessing violence potential in early and later adolescence.
The participants in this study were 415 adolescents incarcerated at a county juvenile detention facility or attending a large Christian-based high school. The Life Challenges Questionnaire-Teen Form was administered to a random sample of 110 adolescents incarcerated at a midwestern county juvenile justice complex and to 305 adolescents attending a large, private, Christian-based high school located in an urban setting in the Midwest. The two groups of adolescents range in age from 11 to 18. Each youth took the paper and pencil questionnaire under the supervision and direction of a test administrator. All individuals present on the day of data collection were offered the opportunity to participate at each of the sites. The questionnaire was administered to 110 adolescent males and females at the juvenile detention facility and 305 adolescent males and females at the high school. Of the 110 individuals in the detention sample, 11 withdrew prior to completing the questionnaire. All 305 students in the school sample completed the questionnaire. The final sample included 99 detainees and 305 students.
The detainees were incarcerated at a midwestern county juvenile detention center. Lengths of incarceration at the detention facility tend to be rather brief, as most individuals are pre-trial detainees. The average stay is two to three weeks. Adolescents are typically detained at the facility if there is a high risk of reoffending or fleeing, if crimes are particularly serious or violent, or if the minor has violated probation and is sentenced to the requisite 30-day incarceration period stipulated for such offenses. The types of crimes vary considerably among the detainees from repeated status offenses to aggravated battery, aggravated sexual assault, and murder. The children and adolescents incarcerated at the detention facility often have serious emotional difficulties and long histories of physical and sexual abuse as well as neglect. They often come from highly chaotic family environments with little structure, permissive parenting, inconsistent limits, and poor boundaries between parental and spousal subsystems. The families of the adolescents often have long-standing parental drug and alcohol abuse, mental illness, divorce, and relocations. Most of the minors are from minority populations, particularly of Latin or African American descent. In addition, many come from low-income or working class families, with either single or extended family parenting. Many of the minors are wards of the Department of Child and Family Services (DCFS) living in either group homes or foster care.
The non-detention participants are freshman, sophomore, junior, and senior girls and boys at a private Christian-based high school in a nearby western suburb of a major urban city. The student body is comprised of slightly more boys than girls. Of the 305 students, 40% are African American, 40% are Caucasian, 12% are Hispanic, and 2% identify with some other race or ethnicity. All students live within walking distance of the school. However, the neighborhoods surrounding the school are fairly segregated and vary widely with regard to average income. The minority students tend to live in one of the poorest suburban communities. The Caucasian students come from more affluent areas that might be considered upper-middle class, with a number of them living in upper-class neighborhoods. The Hispanic population is relatively new to the area but tends to live in low-income housing similar to the African American population and attend the high school sampled in the present investigation. A significant proportion of the students are expected to attend college based on the high school's college placement statistics, with more than 90% of graduates going on to college. Students that do not go to college often enter the armed services. The students of this Christian high school are held to a higher code of conduct than might be expected in the public school environment and few fights or violent altercations occur between students. However, the campus is open and individuals from the neighboring public schools will, on occasion, incite violent confrontations with students attending the school sampled in this study.
The Life Challenges Questionnaire-Teen Form (LCQ-TF) includes a total of 120 items. Of these, 28 items ask respondents to rate the degree to which particular statements apply to their life, with Likert-type scales including the response choices "always true," "sometimes true," and "never true." Examples include statements such as "I feel pressure to join a gang," "I think my parents/guardians do their best to take care of me." "I have a place to go where I can be alone at home," and "I get angry easily." The questionnaire also includes 29 forced-choice items asking respondents to answer either yes or no to particular questions about their lives. Examples include, "Do you belong to a club at school or in your community?" "Are there students in your school who have guns or other weapons?" "Do your parents or guardians yell at you a lot?" and "Have you been in a fight within the last year?" The next set of 29 items asks respondents about various demographic and family structure variables as well as questions on academic performance, school connectedness, involvement in various recreational activities, substance abuse, and aggressive behaviors. This section includes questions about the respondent's neighborhood and family environment, including the family make-up (nuclear, extended, stepparent, single-parent, number of siblings) the number of people living in the home, the nature of rules and discipline within the home, and the type of neighborhood where the respondent lives (rural, suburban, or urban). Other questions focus on extracurricular activities such as volunteer work, television, computer, and video game use, and time spent with family members or friends. The following section includes 16 forced-choice true or false questions about substance abuse within the home and by peers, history of contacts with the justice system, attitudes toward the police and neighborhood, and accessibility and use of guns and other weapons. Examples include, "I have shot a real gun." "I could never be friends with a police officer" and "Nobody at my house uses drugs." The final set of 24 questions asks the respondent to rate the degree to which certain statements apply to them using a Likert-type scale with responses of agree, disagree, and somewhat agree. This final section relates to the respondents self-perception including items such as, "People think I am dumb, but I am smarter than they think." "I don't think I will get to go to college." "I have many friends." and "I am hard to get along with most of the time." Throughout the inventory similar questions are repeated in different formats to provide a measure of the respondents' truthfulness in answering the questions, as well as the level of investment in carefully reading and responding to them.
The Risk Assessment Index (RAI). While all items in the LCQ-TF reflect known correlates of youth violence, 53 critical items are scattered throughout the questionnaire which make up the Risk Assessment Index. Each of the 53 items has one response, considered the critical endorsement, which previous research has established as a particularly strong correlate of youth violence and aggression. Each critical endorsement is assigned a point value of 1, while all other responses receive 0 points. Overall scores are computed by totaling the number of critical endorsements the respondent has made. This provides a composite Risk Assessment Index (RAI) score ranging from 0-53. Thus, the RAI provides an overall estimate of risk for youth violence based on the respondents' endorsement of known predictors. The higher the RAI score the greater the risk for problems in coping and involvement in violence. The specific items that make up the Risk Assessment Index are included in Appendix 1.
Since the purpose of the present investigation is to assess the potential utility of the Risk Assessment Index as an effective, brief screening tool to identify young people at risk for violence, a comparison is made of a sample of adolescents in the juvenile detention population with the adolescents attending a private high school.
Christian high school. After fulfilling requirements for informed consent and protection of human subjects, all students present on the day of data collection were offered the opportunity to participate in the study. All 305 students in attendance completed the questionnaire. The researchers tested participants in groups of 20, using the paper-and-pencil version of the Life Challenges Questionnaire-Teen Form (LCQ-TF). The examiners assigned each participant to a desk within his or her homeroom class at the school, where the self-report survey was opened to the instruction page. The researchers previously assigned each individual a code number, which was pre-printed on the front page of the questionnaire. The examiners asked participants to begin the survey only when instructed to do so. Prior to administration of the LCQ-TF, the researchers informed the participants that they had been selected to participate in a study about teenagers. They were told that their participation was voluntary and that they were allowed to withdraw at any time. The examiners then informed them that the results of the study would add to the current body of knowledge about life as a teenager but would have little direct benefit to them. Subsequently, the examiner instructed the participants to answer each question as it pertained to them and to answer as truthfully as possible. The examiner explained that individual responses would remain confidential and anonymous and that no identifying information would be included in any publication of the results. The examiner also informed the participants that their individual responses would never be made part of their school record or shared with their parents. The examiners provided the participants with verbal instructions that were the same as those appearing on the first page of the questionnaire. The researchers explained the basic structure and format of the LCQ-TF. The examiners informed the participants that the questionnaire included both multiple choice and true-false items. The examiners also instructed the participants to answer all questions to the best of their ability, leaving no blank responses. The examiners then informed the participants that the questionnaire was expected to take 20 minutes to complete but that there was no time limit. Finally, the researcher reminded participants that the test administrator was available throughout testing should they need help. The researchers then instructed participants to begin the questionnaire. At the conclusion of the test, the researchers thanked the participants and asked them to remain seated at their workstations until dismissed. After all participants completed the survey, the researchers gave them the opportunity to ask any questions and fully debriefed them as necessary.
Juvenile detention facility. Procedures at the detention facility followed those at the high school as closely as possible. The researchers offered all individuals present in juvenile detention the opportunity to participate. Of the 110 detainees incarcerated on the day of testing, 11 withdrew from the study prior to completing the survey. Testing occurred during the detainees' designated school time, between the hours of 9 A.M. and 2 P.M. The researchers tested participants within their home classrooms, in groups of 10 to 12. They were seated at individual desks around the room to prevent sharing of responses and cross talking. The examiners reminded the detainees that their participation was voluntary and that they had the right to withdraw at any time. Confidentiality of their responses to test items was assured. The adolescents in the detention facility received the same set of instructions as outlined in the above procedure for the adolescents at the private high school. The examiners then instructed participants to begin the questionnaire whenever they felt ready to do so. When all participants were finished with the questionnaire, the researchers gave them the opportunity to ask any questions and fully debriefed them as needed.
The primary objective of this analysis is to indicate whether the Life Challenges Questionnaire-Teen Form is successful in differentiating between a sample of adolescents with a known criminal/violent history (detention group) and a sample of adolescents without such a history (school group), thus lending initial support for its validity as a risk-assessment screening tool. The analysis consists of two parts. First, a comparison is made of the two groups of adolescents in terms of describing risk for violence on the basis of responses to the Life Challenges Questionnaire, which examines four areas of life challenges: (1) peer relationships, (2) school and community environmental experiences, (3) family factors, and (4) personality and behavioral factors. Differences are tested for significance by means of the chi square statistic. Secondly, the analysis will turn to an examination of risk assessment based on the Risk Assessment Index (RAI) which was created to measure adolescents' risk for violence. This part of the analysis is guided by the following hypotheses formulated on the basis of previous research:
1. Significantly higher overall RAI scores were expected in the detention sample compared to RAI scores in the school sample.
2. Girls in both groups were expected to score significantly lower than boys on the RAI.
3. Caucasians were expected to score significantly lower on the RAI when compared to minority adolescents.
4. Adolescents between the ages of 15 and 18 were expected to score significantly higher on the RAI than their 12- to 14-year-old counterparts.
These hypotheses are tested by means of one-way analysis of variance (ANOVA) and multiple regression is employed to assess the relative impact of age, grade level, race, and sex on risk for violence as measured by the Risk Assessment Index.
Descriptive Analysis Based on the Life Challenges Questionnaire
The detention and non-detention samples differed significantly in their responses to 79 of the 120 questionnaire items on the Life Challenges Questionnaire. Although the actual item-by-item frequencies and percentages are not presented here for consideration of space, the results reveal a pattern indicating that detainees were more likely than the high school students to endorse items associated with risk factors for violence. These results are summarized below:
Adolescents in the detention group were more likely to come from single-parent, step-parent, or non-biological living arrangements and large families (4 or more children). The detainees were also more likely to witness physical aggression between their parents or guardians, disobey their curfew, and to not be caught or disciplined by their parents when they did something wrong. Parental disengagement was also more common in the detention sample. While detainees always had somewhere to go to be alone at home, non-detainees were more likely to have privacy in their homes intermittently. Finally, detainees were less likely to go on vacation with their families and to have faith in their parents' ability to take care of them and meet their needs.
Detention group members were more likely to associate with gangs, be a member of a gang, or experience pressure to join a gang. Along similar lines, peer groups that foster and support interpersonal aggression, violence, and retribution were more common among adolescents in the detention group. In addition, detainees were more likely to associate with substance-abusing peers and to be accepting of peers who abused substances and indicated possession and use of guns. Detainees spent significantly more time with their friends on a daily basis and were more likely to have substantially older friends. Lastly, detention group members experienced more frequent bullying and teasing from peers.
Community and School Environment
Adolescents in the detention sample were more likely to receive below average and failing grades in their coursework and to experience parental dissatisfaction with their academic performance. Investment in academic success was also lacking, with fewer detainees completing their homework assignments on a regular basis. Detainees were also more likely to lack other positive school-related achievements, such as student leadership and involvement in school-connected clubs and extracurricular activities. However, detainees were more likely to experience success in athletics or the arts rather than academics. Detainees were suspended or expelled from school more frequently, indicating greater conflict with teachers and other authority figures. In addition, adolescents in the detention sample were more likely to attend schools and live in communities with unsafe environments. Lastly, detention-group adolescents were less likely to engage in volunteer work or participate in community-based clubs or extracurricular activities, suggesting weak ties and bonds to the community.
Personality, Behavioral, and Psychological Factors
Adolescents in the detention group were more likely to feel overwhelmed by their problems, to feel alone in the world, and to have been hurt badly by someone they love or care about. Poor self-esteem was also more common among detention-group adolescents, with detainees reporting less satisfaction with their physical appearance and less confidence in their own abilities. Detainees were more likely to see aggression and revenge as an acceptable means of solving conflicts and problems and to experience anger and frustration more readily than their non-detention student group counterparts. Behavior problems, authority conflicts, and trouble with the police were more common among adolescents in the detention group. Antisocial behaviors, such as harming animals, stealing and theft, and physical aggression occurred more frequently in the detention sample. Detainees began using alcohol, other drugs, and cigarettes at an earlier age and were more likely to use them throughout adolescence. They were also more likely to have tatoos and body piercing and to enjoy more violent video games than the non-detention students.
Based on an item-by-item descriptive analysis, these results clearly indicate that the Life Challenges Questionnaire differentiates between the responses of two groups of adolescents, one with a known criminal/ violent history and a comparison sample of students from a private, religious high school. We turn now to a specific focus on the Risk Assessment Index as a scale to assess risk for adolescent violence.
Analysis of Adolescent Risk: The Risk Assessment Index
The 53 items that make up the Risk Assessment Index (RAI) and the critical response category for each item included in the scale, along with a comparison of responses in the scoreable direction by the two groups, are included in Appendix 1. A reliability analysis of the Risk Assessment Index was performed producing a reliability coefficient (alpha) of .87, which means that this scale demonstrates good reliability.
The results of the analysis of variance to test the specific hypotheses are presented in Table 1. As predicted in hypothesis 1, adolescents in detention reveal higher scores on the Risk Assessment Index than their non-detention counterparts. The mean RAI score for the detention sample (M = 17.56) is significantly greater than the mean RAI score in the non-detention sample (M = 11.58), F = 49.06, p < 0.001. Hypothesis 2 was also supported in that males (M = 14.26) received significantly higher RAI scores than did females (M = 11.14), F = 14.66, p < 0.001. However, hypothesis 3 received mixed results. Whites were expected to have lower RAI scores than minority group adolescents. Although whites received significantly lower RAI scores (M = 11.53) than African Americans (M = 13.13), Hispanic/Latinos (M = 16.17), Native-Americans (M = 19), Asian Americans (M = 14.00), and individuals of Arabic descent (M = 25), contrary to expectations, biracial individuals received lower RAI scores (M = 10.94) than whites. It should also be noted that two of the racial-ethnic group categories included only one individual, thus, making interpretations based on these group comparisons less meaningful. Finally, hypothesis 4 did not receive support. Contrary to expectations, adolescents between the ages of 15 and 18 did not score significantly higher on the Risk Assessment Index than their 12- to 14-year-old counterparts. Multiple regression analysis (Table 2) further revealed that in relative terms (i.e., when controlling for the other predictors), detention versus non-detention group status is the strongest predictor of the Risk Assessment Index (RAI) score ([beta] = .310, p < 0.001). Other significant predictors include being male ([beta] = .131, p < 0.05), African American ([beta] = .137, p < 0.05) or Hispanic ([beta] = .157, p < 0.05).
DISCUSSION AND CONCLUSION
The present investigation sought initial validation for the Life Challenges Questionnaire-Teen Form (LCQ-TF) as a brief risk assessment tool. The LCQ-TF assesses risk factors for adolescent violence previously established in the research literature, including negative peer influences, gang involvement, poor academic performance and disengagement from school, disruptive family and community environments, substance abuse, low self-esteem, trauma and physical abuse, and negative personality and psychological factors. The detention and non-detention student samples responded differently to 79 of the 120 items on the questionnaire. With regard to the family environmental risk factors, family chaos and disruption, disengagement among family members, a strong reliance on rules, and a lack of consistent discipline or enforcement of these rules were more likely to be present in the detention group. With regard to peer relationships, the detention group was more likely to identify with gangs, display positive attitudes toward aggression and violence, greater use of substances, and toward increased availability and use of guns. Among school and community risk factors, detention group members were more likely to display poor academic performance, limited school-related achievements, conflict with teachers and administrators, less involvement in organized extracurricular activities, weak ties to the community, and to perceive their school and community environments as unsafe. With regard to personality, behavioral, and psychological characteristics that serve as significant risk factors, the detention group was more likely to display poor coping skills, antisocial behaviors, consistent defiance of authority, poor self-esteem, loneliness and isolation, and low frustration tolerance. This analysis, therefore, supports the validity of items in the Life Challenges Questionnaire (LCQ) in terms of being able to differentiate between a group of adolescents with a known history of delinquency (detention sample) and an adolescent group with no such known history (non-detention sample) covering these four main areas of adolescent risk--the family, peer relationships, the school and community environments, and personality and psychological factors.
Further analysis of adolescent risk based on LCQ items which composed a risk assessment scale (The Risk Assessment Index-RAI), provided additional support for the utility of the LCQ as a screening tool for assessing adolescent risk for violence. The RAI demonstrates good reliability as measured by Cronbach's alpha and scores obtained on the RAI indicate that it captures and accurately assesses those risk factors most strongly correlated with youth violence, thus providing a valid indicator of its potential as an early prevention screening instrument. As predicted, adolescents in the detention group scored significantly higher on the Risk Assessment Index than those in the non-detention student group. Moreover, the results indicate that detention group status is the single strongest predictor of the RAI score. In addition, males and minority populations, specifically African Americans and Hispanics, score higher on the Risk Assessment Index. These findings lend support to the validity of the LCQ-TF and the RAI, more specifically, and are consistent with previous research on risk factors for youth violence and acting out.
Limitations and Future Directions
Despite indicating the potential utility of the LCQ-TF as a screening tool, limitations in the study's design and findings should be noted. The study included relatively small and restricted samples. The detention sample included few female adolescents and some of the racial-ethnic categories and age groups included few individuals, overall. Moreover, the two samples included adolescents from restricted populations, namely a private Christian high school in an urban setting and a suburban county juvenile center. These two groups may have unique characteristics and experiences that are not representative of the general population of adolescents. Future validation analysis should include a broader range of adolescents with regard to age, race, family background, socioeconomic status, gender and other characteristics to increase the generalizability of results. In addition, the present investigation did not control for such factors as socioeconomic status, type of criminal history, or the type of testing environment. Future studies controlling for these variables might provide greater information with regard to the strengths and limitations of the LCQ-TF.
Despite these limitations, the present investigation provides evidence which lends initial support for the validity of the Life Challenges Questionnaire-Teen Form as an effective risk-assessment screening tool for identifying youth at risk for poor coping and violence. Its relatively simple and straightforward design and scoring procedures make it a useful tool for clinicians, parents, and educators. Early identification and intervention are perhaps the most important factors in the prevention of youth violence. Once identified by the LCQ-TF, those at high risk can be referred to trained mental health professionals for more comprehensive assessment and to psychological programs before violence and other negative behavior become a serious problem.
Appendix 1 Selection of Critically Endorsed Responses on the 53 Items That Make Up the Risk Assessment Index by Adolescent Group Status (Detention vs. Non-Detention) Critical Non- RAI Item Response Detention Q1. People think I'm a Always true 2.3% gang member Q2. I feel pressure by Always true 1.0% others to join a gang Q4. Parents do best to Always true 85.5% take care of me Q10. My parents hit me Always true 1.3% for no reason Q11. I see my parents Always true 0.7% strike each other Q12. I get angry easily Always true 16.5% Q13. When someone Always true 12.2% makes me angry, I get even Q14. If you mess with Always true 16.1% me/friends, you will get hurt Q17. I feel safe at school Never true 6.9% Q19. I don't get caught Always true 9.2% when I break rules at school Q20. I wish I did not have Always true 30.2% to go to school Q25. The more violence in Always true 17.0% a video game, the more I enjoy it Q27. Usually I do not Always true 7.9% seem to fit in Q28. I don't mind my friends Always true 19.0% using drugs around me Q33. Been suspended from Yes 34.1% school before Q34. Students in your school Yes 46.2% with guns/other weapons Q35. Wish to be member Yes 5.9% of a gang/crew Q36. Been asked/tempted Yes 25.2% to become gang member Q37. How long have you Recently 0.7% been in a gang Q38. Ever witnessed Yes 61.6% anyone in your family fighting and hitting Q42. Feel like you have Yes 45.4% too many problems to deal with right now Q43. Parent/guardian been Yes 20.3% very ill within past year Q44. Anybody ever hurt Yes 29.2% you badly Q46. Has an adult touched Yes 8.2% you in a way to make you feel uncomfortable Q47. Has there been a death Yes 43.6% in your family within past year Q48. Have you ever been Yes 31.1% so upset that you felt like hurting yourself Q49. Do you ever get so Yes 64.9% angry that you think about hurting someone Q52. Been so angry with Yes 54.1% someone at school that if something happened to them you would not care Q53. Ever killed an animal Yes 8.2% because curious to see what it would feel like Q68. Who do you trust No one 13.5% most of all Q71. People in your school You are ugly 3.6% probably think you are (attractiveness) Q72. What is your major None 7.2% accomplishment at school Q73. Not counting time in None 16.6% school, how much time do you spend with friends every day Q75. How much time do More than 20 4.3% you spend on the Internet hours each week Q76. How much time do you More than 20 18.5% spend watching TV each week hours Q77. How much time do More than 20 4.3% you spend playing video hours games each week Q78. How many physical 4 or more 8.6% fights have you been in within past year Q79. How old were Less than 12 4.7% you when you started years of age to smoke cigarettes/marijuana 12 to 14 years 11.0% 15 to 17 years 14.3% Q80. How old were you Less than 12 6.0% when you first started to years of age drink alcohol 12 to 14 years 18.3% 15 to 17 years 13.3% Q89. Someone has shot a False 85.6% real gun at me Q90. I have NOT held a False 56.2% gun before Q91. I have shot a real True 33.8% gun before Q92. I know how to get a True 45.8% gun if I need one Q93. I think a gun may be True 31.8% in my home Q94. There are kids in my True 49.0% school who have guns Q95. I have friends who True 46.2% carry weapons sometimes Q96. I know kids who are True 70.2% in gangs Q98. Nobody really Agree 28.2% understands me Q101. I am liked at school I am not liked 4.4% because I am smart at school Q102. I think I am I don't care 3.7% healthy and in good about my health shape Q110. I have many friends I have no 2.0% friends Q117. There is someone in Agree 13.6% my home who drinks too much alcohol Q118. Nobody really cares Agree 5.4% about me RAI Item Detention Sig. * Q1. People think I'm a 35.4% 0.001 gang member Q2. I feel pressure by 4.0% 0.01 others to join a gang Q4. Parents do best to 79.6% 0.001 take care of me Q10. My parents hit me 3.0% NS for no reason Q11. I see my parents 3.0% 0.01 strike each other Q12. I get angry easily 40.4% 0.001 Q13. When someone 22.2% 0.05 makes me angry, I get even Q14. If you mess with 31.3% 0.01 me/friends, you will get hurt Q17. I feel safe at school 14.1% NS Q19. I don't get caught 12.1% NS when I break rules at school Q20. I wish I did not have 26.3% 0.001 to go to school Q25. The more violence in 31.3% 0.01 a video game, the more I enjoy it Q27. Usually I do not 5.1% 0.01 seem to fit in Q28. I don't mind my friends 33.3% 0.001 using drugs around me Q33. Been suspended from 88.9% 0.001 school before Q34. Students in your school 61.6% 0.01 with guns/other weapons Q35. Wish to be member 18.4% 0.001 of a gang/crew Q36. Been asked/tempted 56.6% 0.001 to become gang member Q37. How long have you 12.1% 0.001 been in a gang Q38. Ever witnessed 54.5% NS anyone in your family fighting and hitting Q42. Feel like you have 67.7% 0.001 too many problems to deal with right now Q43. Parent/guardian been 29.3% NS very ill within past year Q44. Anybody ever hurt 42.4% 0.05 you badly Q46. Has an adult touched 11.1% NS you in a way to make you feel uncomfortable Q47. Has there been a death 36.4% NS in your family within past year Q48. Have you ever been 24.2% NS so upset that you felt like hurting yourself Q49. Do you ever get so 61.6% NS angry that you think about hurting someone Q52. Been so angry with 58.6% NS someone at school that if something happened to them you would not care Q53. Ever killed an animal 17.2% 0.05 because curious to see what it would feel like Q68. Who do you trust 31.6% 0.001 most of all Q71. People in your school 2.0% NS probably think you are (attractiveness) Q72. What is your major 22.2% NS accomplishment at school Q73. Not counting time in 6.1% 0.001 school, how much time do you spend with friends every day Q75. How much time do 0.0% 0.001 you spend on the Internet each week Q76. How much time do you 12.1% 0.001 spend watching TV each week Q77. How much time do 3.0% NS you spend playing video games each week Q78. How many physical 36.4% 0.001 fights have you been in within past year Q79. How old were 27.3% 0.001 you when you started to smoke cigarettes/marijuana 48.5% 6.1% Q80. How old were you 17.2% 0.001 when you first started to drink alcohol 39.4% 12.1% Q89. Someone has shot a 49.5% 0.001 real gun at me Q90. I have NOT held a 70.7% 0.05 gun before Q91. I have shot a real 62.6% 0.001 gun before Q92. I know how to get a 58.6% 0.05 gun if I need one Q93. I think a gun may be 14.1% 0.001 in my home Q94. There are kids in my 58.6% 0.05 school who have guns Q95. I have friends who 56.6% NS carry weapons sometimes Q96. I know kids who are 87.9% 0.001 in gangs Q98. Nobody really 26.3% NS understands me Q101. I am liked at school 10.1% NS because I am smart Q102. I think I am 5.1% NS healthy and in good shape Q110. I have many friends 4.0% NS Q117. There is someone in 15.2% NS my home who drinks too much alcohol Q118. Nobody really cares 5.1% NS about me * Based on [chi square] test of statistical significance.
Astor, A. R. (1995). School violence: A blueprint for elementary school interventions. Social Work in Education, 17(2), 101-116.
Bischof, P. G., Stith, M. S., & Whitney, L. M. (1995). Family environments of adolescent sex offenders and other juvenile delinquents. Adolescence, 30(117), 157-170.
Bonny, E. A., Britto, T. M., Klostermann, K. B., Hornung, W. R., & Slap, B. G. (2000). School disconnectedness: Identifying adolescents at risk. Pediatrics, 106, 1017-1031.
Book, S. A., & Knap, A. M. (2001). Criterion validity of the Holden Psychological Screening Inventory Social Symptomatology Scale in a prison sample. Psychological Assessment, 13(2), 249-25.
Dorman, M. S., & Pealer, N. L. (1999). Youth 97: Summary data from the Youth Risk Behavior Survey. Journal of School Health, 69(3), 128-129.
Dusenbury, L., Falco, M., & Lake, A. (1997). Nine critical elements of promising violence prevention programs. The Journal of School Health, 67, 409-415.
Ellickson, L. P., & McGuigan, S. K. (2000). Early predictors of adolescent violence. American Journal of Public Health, 90(4), 566-572.
Farrell, D. A., & Sullivan, N. T. (2000). Structure of the Weinberger Adjustment Inventory Self-Restraint Scale and its relation to problem behaviors in adolescence. Psychological Assessment, 12(4), 294-401.
Flannery, J. D., Singer, I. M., & Webster, K. (2001). Psychopathic traits and conduct problems in community and clinical referred samples of children: Further development of the Psychopathy Screening Device. Psychological Assessment, 12(4), 382-393.
Guetzloe, E. (1999). Violence in children and adolescents--a threat to public health and safety: A paradigm of prevention. Preventing School Failure, 44(1), 21-25.
Kelley, T. B., Huizinga, D., Thornberry, P. T., & Loeber, R. (1997). Epidemiology of serious violence. The Office of Juvenile Justice and Delinquency Prevention Youth Development Series, 1-20.
MacArthur Research Network on Mental Health and the Law. (2001). The MacArthur violence risk assessment study: Executive summary. MacArthur Research Network website, http://macarthur.virginia.edu/risk. html.
Office of Juvenile Justice Delinquency and Prevention. (1999). Juvenile offenders. Juvenile Offenders and Victims: 1999 National Report, Chap. 3.
Rachuba, L., Stanton, B., & Howard, D. (1995). Violent crime in the United States: An epidemiologic profile. Archives of Pediatrics & Adolescent Medicine, 149(9), 953-960.
Rodney, E. H., Tachia, R. H., & Rodney, W. L. (1999). The home environment and delinquency: A study of African American adolescents. Families in Society: The Journal of Contemporary Human Services, 80(6), 551-664.
Reprint requests should be sent to Marvin P. Dawkins, Ph.D., Professor of Sociology, University of Miami, 5202 University Drive, Merrick Building, Room 121-G, P.O. Box 248162, Coral Gables, Florida 33124-2030.
Table 1 Effects of Group Status (School vs. Detention), Age, Sex, Race, and Grade Level on Adolescent Risks as Measured by the Risk Assessment Index (RAI) One-Way Standard ANOVA Significance Variable n Mean Deviation F Level Group Status Non-Detention 246 11.58 6.58 49.06 0.001 Detention 87 17.56 7.56 Age 11 2 24.00 1.41 1.49 NS 12 1 13.00 -- 13 5 20.80 13.68 14 40 13.25 7.44 15 97 13.02 7.03 16 98 12.81 7.46 17 63 12.76 7.18 18 24 12.67 6.30 Sex Male 192 14.26 7.21 14.696 0.001 Female 130 11.14 7.11 Race Caucasian 120 11.53 6.75 3.744 0.001 African American 119 13.13 6.41 Hispanic/Latino 64 16.17 9.11 Native/American 1 19.00 -- Indian Asian/Asian 2 14.00 12.73 American Bi-Racial/Mixed 17 10.94 6.60 Arabic 1 25.00 -- Grade Level Freshman 84 13.55 7.53 2.168 NS Sophomore 81 13.31 7.59 Junior 73 10.97 6.34 Senior 56 13.39 6.84 Table 2 Multiple Regression Analysis: The Relative Effects of Five Predictors (Group Status, Gender, Race, Age, and Class Standing) on Risk Behavior as Measured by the Risk Assessment Index (RAI) Standardized Regression Predictors Coefficient Beta Detention .310 ** Male .131 * Age -.062 African American .137 * Hispanic .157 * Bi-Racial .030 Sophomore .066 Junior -.062 Senior .134 [R.sup.2] = .195, Adjusted [R.sup.2] = .170. * p < 0.05, ** p < 0.001
|Printer friendly Cite/link Email Feedback|
|Author:||Grinberg, Ilyse; Dawkins, Marva; Dawkins, Marvin P.; Fullilove, Constance|
|Date:||Sep 22, 2005|
|Previous Article:||The role of personality factors in predicting the reported physical health symptoms of Turkish college students.|
|Next Article:||The relation of facial affect recognition and empathy to delinquency in youth offenders.|