A Review of Selected Questionnaires
Community Exposure to Violence Questionnaires
There has been a tremendous increase in the development of new questionnaires. Compared with 10 years ago, there are now numerous questionnaires that measure various forms of child victimization, ranging from general questionnaires that measure exposure to community violence to specialized questionnaires that focus on specific types of victimization, such as child maltreatment or bullying. Tables 1–6 present representative questionnaires for the major forms of victimization. Researchers interested in using a questionnaire should contact the author or publisher and request permission to do so. Some questionnaires are copyrighted. Contact information for the authors of the questionnaires reviewed here is available in the articles cited in the tables. Web site addresses for two questionnaires are also provided in the tables.
Community Exposure to
The most general questionnaires are those that ask about exposure to violence in a child’s community. These questionnaires were developed in part because of increased concern about and awareness of children’s exposure to street crime. Thus, often the questions concentrate on different forms of physical assault and, typically, at least half of the questions ask whether a child has witnessed violence as opposed to directly experiencing violence. Because of their focus on street crime, many of these measures also contain questions about whether a child has been exposed to drug deals or witnessed arrests, both of which, strictly speaking, fall outside most definitions of violence. Many of these questionnaires have been used primarily with samples of high-risk children who live in areas that are known to have above-average crime rates.
These questionnaires have established that a very large percentage of children have, in fact, experienced and/or witnessed violence. Witnessing assault, in particular, is typically reported by a clear majority of children, with exposure rates often in the 85-percent range (e.g., Hill and Jones, 1997). Most community violence questionnaires have shown good internal consistency and correlate with measures of depression, posttraumatic stress symptoms, and behavioral problems. Their primary disadvantages are the inadequate attention to violence by family and intimate perpetrators, lack of national norms, and overestimation of children’s ability to understand terms such as “robbery.”
The most widely used community violence questionnaires are those developed by Richters, Martinez, and Valla (1990), Things I Have Seen and Heard (designed for first and second graders), and Richters and Martinez (1993), Survey of Children’s Exposure to Violence (SCEV; designed for older children). The My Exposure to Violence (My ETV) questionnaire (Selner-O’Hagan et al., 1998) is a fairly new instrument that has been developed with careful attention to reliability and validity. For very young children, the Violence Exposure Scale for Children (VEX) (Fox and Leavitt, 1995) has translated the questions typical of this class of questionnaires into cartoon-like drawings. For information on these and other community violence questionnaires, see table 1.
Child maltreatment questionnaires focus on the types of victimization that are investigated and treated by CPS agencies. These include physical abuse, sexual abuse, and neglect, although not all scales cover all three forms of maltreatment. Assaults by caregivers are usually the focus of these questionnaires, but sexual assaults by noncaregivers may be included because these are also sometimes investigated by institutions investigating child maltreatment.
Initially, there was some skepticism about whether caregivers would report their own use of violence against their children, but a number of national studies (e.g., Straus and Gelles, 1990; Straus et al., 1998) have established that caregiver self-reports produce higher estimates of child maltreatment than official data from CPS agencies, physicians, or police. There is some evidence, in fact, that parents will report more incidents of parent-to-child violence than children will (e.g., Grych, 1998; Jouriles and Norwood, 1995; Kruttschnitt and Dornfeld, 1992). As with community exposure measures, the reliability and validity of many child maltreatment scales are well-established. The primary problem with most child maltreatment scales is one of interpretation. Because spanking and other forms of physical discipline are legal, it can be difficult to determine which questions clearly indicate physical maltreatment. Neglect, which often is characterized by a pattern rather than by a single episode of inattention, has also proven to be hard to identify through self-reports. Sexual abuse is generally easier to identify because virtually any sexual contact between an adult and a child is considered abusive, but even that can be harder to distinguish with older adolescents who may be engaging in what they consider to be consensual acts.
The most commonly used measure of physical assault by parents is the Conflict Tactics Scales, which has recently been revised to include questions on neglect and sexual abuse (Straus et al., 1998). This measure was originally designed as a research tool, however, and its physical and psychological aggression questions do not map readily onto CPS categories. The Child Abuse Potential Inventory (Milner, 1986) is another very commonly used tool, but it focuses on stress and other correlates of abuse, rather than measuring maltreatment directly. Most other measures of child maltreatment are designed to gather information from adults about their own childhood histories, although some of these have been used with teens (see table 2).Sexual Assault Questionnaires
Sexual assault questionnaires focus in depth on the many forms of sexual victimization, which include not only rape but also fondling, sexual harassment, and exposure to or involvement in pornography. Furthermore, most sexual assault questionnaires ask about a broad spectrum of coercive sexual behaviors, which can include the use of alcohol or drugs, threats to end a relationship, and physical assault and weapons to force a juvenile into sexual activity. These measures have shown that rates of sexual assault are much higher than previously thought and that it is important to avoid the use of emotionally laden and stigmatized terms like “rape” when seeking self-reports. The main difficulties with measuring sexual assault among children are the sensitive nature of the questions and the difficulty of asking about sexual experiences in a way that young children will understand.
There are few questionnaires that are designed to elicit information about children’s current experiences of sexual assault (see table 3). The questionnaire that is most commonly used with college-age youth and adults is the Sexual Experiences Survey (SES) (Koss and Oros, 1982). It has also been adapted for use with girls of high school age (Gidycz and Koss, 1989). Finkelhor’s (1979) series of questions have been used in many studies but primarily to obtain reports from adults about past experiences of sexual assault. Many studies have used one or two questions to obtain sexual assault histories, but it is known that such brief measures will produce lower reported rates (Bolen and Scannapieco, 1999) and therefore should be avoided when possible. Nonetheless, many studies of sexual assault among juveniles have used such brief measures. Examples of such studies include the Washington State Survey of Adolescent Health Behaviors (Stock et al., 1997), which surveyed girls as young as those in eighth grade, and the National Survey of Family Growth (NSFG) (Abma, Driscoll, and Moore, 1998), which surveyed girls as young as age 15.
Peer Victimization Questionnaires
Assaults by peers or siblings are one of the most common forms of violence against children. Peer victimization is increasingly recognized as another important form of victimization and includes bullying, physical assaults, emotional abuse, and sexual offenses. Gang-related assaults are another form of peer assault, but very little information is available about the victims of gang assaults. There is no widely accepted standard questionnaire in the relatively new peer victimization field. The Social Experience Questionnaire—Self Report (Crick and Bigbee, 1998; Crick and Grotpeter, 1996) includes questions not only about physical victimization (for example, “How often do you get hit by another kid at school?”) but also about relational victimization (for example, “How often does another kid say they won’t like you unless you do what they want you to do?”). The Perceptions of Peer Support Scale includes four questions on physical and emotional victimization and has been used with children as young as age 5 (Kochenderfer and Ladd, 1996; Ladd, Kochenderfer, and Coleman, 1997) (see table 4).Witnessing Violence Questionnaires
As mentioned above, most questionnaires on community exposure to violence include questions on witnessing, in addition to experiencing, violence. Some of these questionnaires also contain questions about what is usually called “indirect exposure” to violence, which includes such things as having a close family member assaulted, but not when the child was physically present.
One form of witnessing violence that is currently receiving a lot of attention is the witnessing of domestic violence. The questionnaires concerning these issues focus on assaults between parents (or between a parent and a stepparent, boyfriend, or girlfriend) that occur when children are either physically present or can see or hear the violence. The partner violence version of the Conflict Tactics Scales (Straus et al., 1996) has been adapted to measure children’s witnessing of violence and is the most commonly used measure in this field. It has been used with children as young as age 8 (Margolin and John, 1997) and more commonly with children in sixth grade and above (e.g., DuRant et al., 1995; Kruttschnitt and Dornfeld, 1992; Rogers and Holmbeck, 1997) to collect data on witnessing violence between parents. Other measures of family conflict include the Family Environment Scale (Moos and Moos, 1986) and the Children’s Perception of Interparental Conflict Scale (Grych, Seid, and Fincham, 1992) (see table 5).Multidimensional Questionnaires
The newest trend in the development of juvenile victimization questionnaires is the creation of interviews and surveys that ask about a number of different forms of victimization. These multidimensional questionnaires have some important advantages. One advantage is the ability to learn about the overlap among forms of victimization. Another advantage is that multidimensional questionnaires can do a better job than unidimensional questionnaires of identifying all of the victims in a group. For example, a questionnaire that asks only about sexual abuse will classify as nonvictims those youth who have experienced physical abuse or any other non-sexual victimization. The main disadvantage of this approach is that it requires more questions.
The National Crime Victimization Survey (NCVS) is a well-known example of a multidimensional survey. It has questions on both conventional crime victimization and sexual assault, and the newest version has improved the effort to identify family and acquaintance perpetrators. NCVS has some of the best nationally representative data available. It is a very complicated survey to administer, however, and still tends to obtain reports primarily about stereotypical criminal assaults. It also uses vocabulary and concepts that may be difficult for children to fully understand. The new Childhood Experiences of Violence Questionnaire (Walsh and MacMillan, 1999), in a series of 12 questions (with some followup), includes items on peer victimization, witnessing domestic violence, caregiver physical assault, and sexual assault. It is currently being tested in Canada.
The comprehensive Juvenile Victimization Questionnaire (JVQ) (Hamby and Finkelhor, 1999) covers the broadest range of victimizations. This questionnaire includes sections on conventional crime, child maltreatment, peer victimization, sexual assault, and witnessing and indirect violence. It also includes a section on extraordinary or catastrophic violence, including public shootings, hate crimes, terrorism, and other rare but very serious forms of violence. The sections can be used individually or together to provide a comprehensive overview of victimization. The questionnaire is designed to correspond to conventional categories of crime and child maltreatment. It also is currently being tested in a number of studies. Table 6 provides further information on multidimensional questionnaires.