The present study examines the correlates of indirect and verbal victimization by peers at school. The research is based on a nationally representative sample of 16,604 students in grades 7 through 11 in 324 schools across Israel. Self-administrated anonymous questionnaires were completed during class. Hierarchical linear modeling was used to examine the relationships between students' victimization and student- and school-level variables. The study gives an indication of the importance of making distinctions between verbal victimization and indirect forms of victimization. For example, the findings indicate that boys reported more verbal and less indirect victimization than did girls. Students from schools with higher proportions of families of low socioeconomic status were more indirectly victimized but were less verbally victimized. However, the research also reveals some similarities between the patterns of relationships of the two victimization forms and certain correlates. Implications of understanding the differences between these two victimization types are highlighted.
KEY WORDS: gender; indirect victimization; school ecology; socioeconomic status; verbal victimization
School violence has been a major concern for students, parents, school staff, and the general public. A considerable part of past empirical studies on school violence has focused on direct, overt behavior patterns, such as physical violence (for example, hitting, kicking) or verbal violence (for example, cursing, calling names) (for example, Grunbaum et al., 2002;Lowry, Sleet, Duncan, Powell, & Kolbe, 1995; Olweus, 1993a). Research on the prevalence of direct aggression and victimization by peers at school has demonstrated that boys are significantly more involved in such behaviors than are gifts, either as bullies or as victims (Grunbaum et al., 2002; Olweus, 1993b; Ostrov & Keating, 2004), a finding that has led to a focus on male forms of aggression (Crick, 1997). However, in the past several years studies have demonstrated the importance of the investigation of indirect forms of violence (including relational and social forms) in addition to direct (physical and verbal) violence, which was found to be more characteristic of girls than of boys (for example, Bjorkqvist, Lagerspetz, & Kaukiainen, 1992; Crick, 1997; Osterman et al., 1998; Ostrov & Keating, 2004; Owens, Daly, & Slee, 2005; Tapper & Boulton, 2005).
Despite the slight differences between relational, indirect, and social forms of aggression (see Coyne, Archer, & Eslea, 2006), they all relate to behaviors by which intentional harm is caused to others by damaging their social relationships or feelings of peer acceptance (Crick, Casas, & Mosher, 1997; Crick & Grotpeter, 1995; Ostrov & Keating, 2004); for example, excluding individuals from group activities, spreading rumors, or maliciously gossiping about them. Coyne et al. (2006) compared physical, verbal, and these indirect forms of aggression. They found that when comparing physical and verbal aggressive behaviors with the various forms of indirect behaviors mentioned it is clear the latter are similar enough to each other (and different from the former) to warrant inclusion within one single category. All three forms of aggression contain elements of manipulative and often covert types of aggression. In the present study, we follow the suggestion of Coyne et al. (2006) to decrease confusion and argumentation among researchers of indirect, relational, and social aggression by creating one larger construct, namely indirect victimization that encompasses the behaviors included in these forms of aggression.
In physical victimization, physical damage is the instrument of harm; in verbal victimization, psychological abuse is the instrument of harm. By contrast, relationships serve as the vehicle of harm in indirect victimization (Crick, 1997). Although according to this widely accepted distinction (for example, Bjorkqvist, 1994; Crick, 1997; Ostrov & Keating, 2004) that indirect victimization is different from verbal victimization, some previous research tended to use the dichotomy of direct versus indirect victimization to distinguish between physical and verbal victimization (Bjorkqvist, 1994; Frodi, Macaulay, & Thome, 1977).
The aim of this study is to address separately indirect and verbal forms of victimization by peers, from an ecological multilevel perspective: examining simultaneously individual and school contextual characteristics. We expand on previous studies, many of which have ignored the multilevel nature of the phenomenon, focusing mainly on the child as the unit for analysis. Recently, several school violence studies have examined the contributions of the social and educational contexts as relating to violent behavior (for example, Gottfredson, Gottfredson, Payne, & Gottfredson, 2005; Osher et al., 2004). However, these studies focus on direct verbal and physical victimization. The following section presents a review of the literature on the correlates included in this study.
Boys versus Girls. Studies indicate consistently that boys are more likely to be victimized and to be aggressors of direct physical and verbal violence (Benbenishty, Zeira, & Astor, 2000; Grunbaum et al., 2002; Nansel et al., 2001; Olweus, 1993b; Ostrov & Keating, 2004), whereas girls are more involved in indirect forms of bullying (such as spreading rumors or excluding someone on purpose), either as bullies, victims, or both (Bjorkqvist, 1994; Bjorkqvist et al., 1992; Osterman et al., 1998; Ostrov & Keating, 2004; Owens et al., 2005). Crick and Grotpeter (1995) proposed that when attempting to cause harm to others (that is, aggressing) children do so in ways that are most likely to damage the social goals of the target. Thus, boys are likely to use physical and verbal direct forms of aggression (usually against other boys) that hinder the instrumentally oriented dominance goals that tend to be characteristic of boys (Block, 1983). Girls, however, are more likely than boys to use more covert, indirect aggression (usually against other girls) as a means of achieving such social goals as maintaining status, power, and popularity in peer groups (Crick, 1997; Crick et al., 1997; Leaderbeater, Boone, Sangster, & Mathieson, 2006; Zahn-Waxler & Polanichka, 2004). It should be noted, however, that some studies show that boys are more victimized and more inclined to be aggressors of both direct and indirect forms of victimization than are girls (for example, Baldry, 2004; Leaderbeater et al., 2006). Some studies also show that there are no significant differences between boys and girls in indirect aggression and victimization at school (for example, Tapper & Boulton, 2004).
Grade Level. More cases of physical (both moderate and serious forms), verbal, and indirect victimization by peers at school are reported by younger students than by students in higher grades (Borg, 1999; Nansel et al., 2001; Owens et al., 2005; Scheithauer, Hayer, Peterman, & Jugert, 2006; Smith, Madsen, & Moody, 1999). For instance, Furlong and colleagues (1995), in their study among U.S. secondary school students, found higher levels of victimization in junior high schools (grades 7 to 8) than in high schools (grades 11 to 12). Similarly, Scheithauer et al. (2006) found among German students that self-reported rates of physical, verbal, and indirect forms of victimization were significantly higher for students in grades 5 to 7 than for students in higher grades.
School Climate. Many researchers emphasize the importance of developing positive school climate to reduce school violence, such as having consistent and fair rules, positive student-teacher relationships, and student participation in decision making (for example, Dupper & Meyer-Adams, 2002; Khoury-Kassabri, Benbenishty, Astor, & Zeira, 2004; Payne, Gottfredson, & Gottfredson, 2003; Schreck, Miller, & Gibson, 2003).We have chosen to view the effects of school climate as emanating both from student-and school-level perspectives. In this dual-level approach, we follow Bryk and Raudenbush (2002) who caution against the aggregation bias that may occur when a variable takes on different meanings (and might therefore have different effects) at different organizational levels. School climate, as shown in previous studies (for example, Khoury-Kassabri et al., 2004), may contribute differently to school violence when considered from the perspective of the individual student than when seen as a general school characteristic (that is, the general atmosphere at school).
Socioeconomic Status (SES). The socioeconomic contexts in which children and adolescents grow up significantly correlates with their involvement in violence (Attar, Guerra, & Tolan, 1994; Kupersmidth, Griesler, DeRosier, Patterson, & Davis, 1995; Patterson, Kupersmidth, & Vaden, 1990). Because of privacy considerations, we do not have information on the SES of the individual students. Instead we have aggregated data (that is, the percentage of low-income and undereducated families at school) on the SES of the students' families. Thus the study examines this issue at the school level and explores how students' reports of victimization are associated with the proportion of low SES families at the school.
Ethnic and Cultural Affiliation. Studies indicate that school violence rates differ by ethnicity and culture. In the United States, several studies have demonstrated that victimization and perpetration rates vary by ethnic background. For example, in a study involving a nationally representative sample of about 16,000 U.S. youths (in grades 6 to 10), Nansel et al. (2001) found that Hispanic youths have reported marginally higher involvement in moderate ("bullying others sometimes") physical, verbal, and indirect bullying (12%) and frequent ("once a week or more") bullying (10.4%) compared with African American...