|
...culture country of residence.
These factors follow many levels of hierarchicy--individual students within classes, classes within schools, schools within neighborhoods, and neighborhoods within societies and cultures (see Duncan & Raudenbush, 1999; Lee, 2000). One of the most important research questions concerning school violence is the degree to which these nested ecological factors contribute to student victimization in the school grounds. The current study attempts to provide an answer by examining the effects of students' individual characteristics and school-level variables on students' reports of school victimization.
Duncan and Raudenbush (1999) and others suggest using multi-level approaches to examine differential effects of student, school, and neighborhood characteristics. Lee (2000), for example, emphasizes the importance of the context in which learning occurs, and argues that in addition to the child's personal abilities we should examine the contextual effects on children's behavior. This context may be defined by the children's families, their classmates, the peers with whom they choose to interact, and the teachers who instruct them.
Welsh, Greene, and Jenkins (1999), using a multilevel approach to study school misconduct, argue that research and intervention efforts have too often been piecemeal, examining specific variables and levels of analysis in isolation from one another. Furthermore, these past studies have focused on one level of analysis--either determining who the more violent students are or which are the violent schools. Welsh et al. (1999) simultaneously examined the relative influence of the individual, institutional, and community factors on misconduct at school based on a sample of 7,583 students from 11 junior high schools in Philadelphia. They conceptualized community in two ways--local (the census tract around the school) and imported (aggregated measures from the census tract where the students live). They studied the following predictors of school misconduct: Community poverty and residential stability; community crime, school size, and students' perceptions of school climate; and individual student characteristics. Their results showed that the vast majority of explained variance in student misconduct (94%) was attributable to within-school (individual) factors; school and community-level factors (both local and imported) added only small increments to explained variance (an additional 4.1-4.5%).
These findings contradict various theories emphasizing the roles of the school and community in school misconduct and violence. Although, Welsh et al. (1999) offer a perspective on the relative role of the community in determining levels of school violence and present a challenge to existing theories, the limitations of this important study should not be overlooked. The non-representative sample at the school level included only 11 schools from the Philadelphia School District, with a relatively low student response rate (65%). Another limitation was that the dependent variable "school disorder" was measured using only four questions, restricting the possibility of distinguishing between various manifestations of school violence, such as between severe violence and more moderate physical violence or relational and verbal violence (see also a discussion on this study limitations in Hoffman & Johnson, 2000). Nevertheless, findings of the study highlight the importance of further examining the relative role of community and the school in school victimization.
AN ECOLOGICAL MODEL OF SCHOOL VIOLENCE
The present study examines a multi-level model for the contribution of two levels of victimization to explain the variance in school violence. This study expands research of Welsh et al. (1999) by using a nationally representative sample of students in Israel (162 schools and almost 10,400 students), and by exploring the relationships between several types of school victimization (by serious physical violence, moderate physical violence, threats, and verbal-social violence) and several aspects of the school's ecology.
We adopt Bronfenbrenner's (1979) ecological perspective, which offers a framework integrating the various factors associated with school violence. It describes school violence as an interplay among several relevant subsystems (i.e., student, family, school, and neighborhood). Goldstein (1994) describes this type of nested ecological theory as an interactionist theory that tries to understand human behavior as a "duet" between a person's individual traits and contextual and environmental variables (social and physical). This environment may include other human beings who are involved in the situation in which the behavior occurs (other students, teachers) and includes the physical environment (class and school size, school structure). Along these lines, Astor, Pitner, and Duncan (1996) argue that child development can be understood only within the total context in which the child lives and when development is considered on a variety of levels and in a multitude of natural settings. They use this framework to show how victimization and perpetration in school can result from the interactive effects of subsystems such as students' families, communities, and the larger societal context.
As obvious as this may seem, most of the studies on school violence overlook the multi-level nature of the phenomena, focusing either on students or on schools as units of analysis. The current study aims to expand the knowledge in this field by addressing school violence from a multi-level perspective. It examines the relationships of several types of victimization with student-level factors (gender, school-level, and personal perception of school climate), and schools-level characteristics (school climate, the socio-economic status of students' families and the neighborhood in which the school embedded, and the ethnic affiliation of the school).
We review the predictor variables included in our model and present empirical evidence on their contribution to understanding school violence in subsequent paragraphs.
INDIVIDUAL FACTORS
Grade Level
Younger students have consistently reported greater victimization than older students (Benbenishty, Zeira, & Astor, 2000; Borg, 1999; Boulton & Underwood, 1992; Furlong, Chung, Bates, & Morrison, 1995; Smith, Masden, & Moddy, 1999; Whitney & Smith, 1993); Olweus (1993) found that the proportion of students who were bullied in grades 2-6 (11.6%) was approximately twice as high as in grades 7-9 (5.4%). Furlong et al. (1995) compared students reporting no victimization experiences with students reporting 12 or more victimization experiences in the previous month. There are many more students with high victimization levels in junior high schools (7-8 grades) than in high schools (11-12 grades). In the present study, we hypothesize that students in lower grades will report more victimization than students in higher grades.
Gender Differences
Boys have consistently been found to be more violent than girls (Borg, 1999; Centers for Disease Control and Prevention [CDC], 1996, 2000; Everett & Price, 1995; Fitzpatrick, 1997; Grunbaum et al., 2002; Lowry, Sleet, Duncan, Powell, & Kolbe, 1995; Olweus, 1993). They are also more often victimized by others than girls (Benbenishty et al., 2000; CDC, 2000). Olweus (1993) found that in junior high schools 7.4% of the boys reported being exposed to direct bullying compared with 3.3% of the girls. Similarly, a recent report by Grunbaum and associates (2002) shows that boys were significantly more likely than girls to have been in a physical fight (43.1% vs. 23.9%), and that boys were much more likely to need medical attention because of injuries sustained in a physical fight than girls (5.2% vs. 2.9%).
However, gender gaps are not consistent over the different kinds of victimization. For instance, boys tend to be victimized more often by direct forms of physical violence (e.g., hitting, slapping) and girls are victimized more often by indirect forms of bullying (e.g., rumors, exclusion from groups) (Smith & Sharp, 1994; Sullivan, 2000). Benbenishty et al. (2000) found similar gender differences with regard to victimization by physical violence and by threats, but found much smaller differences in verbal-social victimization (see Nansel et al., 2001 for similar USA conclusions). In this study, we examine the relationship between gender and different types of victimization.
Perception of School Climate
Students' reports of victimization are influenced by their perception of school climate (Astor, Benbenishty, Zeira, & Vinokur, 2002a). This perception of school climate may be indirectly associated with, for example, fear of attending school due to violence. Many researchers emphasize the importance of developing a positive school climate in order to reduce school violence (Colven, Tobin, Beard, Hagan, & Sprague, 1998; Dwyer, Osher, & Hoffman, 2000; Flannery, 1997; Fraser, 1996; Stephens, 1994). Astor et al. (2002a) found that students' judgments of their schools' overall problem with violence were related directly to school climate characteristics.
Schools with policies for dealing with violence that include clear, consistent, and fair rules may be able to reduce violence compared with schools without such a policy (Adams, 2000; Astor, Vargas, Pitner, & Meyer, 1999; Limper, 2000; Olweus, 1991; Smith & Sharp, 1995). Supportive positive relationships with adults may reduce students' alienation from their school and allow them to overcome their emotional and behavioral problems (Dwyer et al., 2000). In addition, student participation in decision making and in designing ways to prevent school violence may be effective. This participation may enhance students' involvement in the school and increase their interest in a peaceful school. Students disengaged from school have little or no investment in acting appropriately (Flannery, 1997). We therefore hypothesize that lower victimization rates will be reported in schools with positive school climate (i.e., clear policy against violence, supportive relationships between teachers and students, and student participation in decision making).
We have chosen here to view the effects of school climate and gender as emanating both from a student-level and from a school-level perspective. In this dual-level approach, we follow Bryk and Raudenbush (1992) who caution against an aggregation bias that may occur when a variable takes on different meaning (and may therefore have different effects) at different organizational levels. For instance, at the student level, low socioeconomic status (SES) may be correlated negatively with academic achievement, but at a school level, schools characterized by below-average SES may have higher academic achievements (for example, they may receive more resources that contribute to higher achievements).
Similarly, we think that school climate may contribute independently to school violence when considered from the perspective of the individual student and when seen as a school characteristic. Thus, for instance, individual students' perception of the extent to which they can participate in decision making in school may not be related to their individual level of victimization. In contrast, at the school level, the proportion of students in school who feel that they participate in decision making may be strongly associated with effective school leadership, responsible among other things, for lower levels of school violence.
Examining both gender and school climate as both student- and school-level variables allows us to ask whether boys in a school tend to report more or less victimization than the girls in their school. We also ask whether schools with a high concentration of boys (or perhaps even boys-only schools) exhibit extremely high levels of violence, perhaps due to unique social dynamics that are characteristics of an all-male peer group.
SCHOOL-LEVEL FACTORS
Community and Family Context
The neighborhoods in which children and adolescents grow up strongly...
NOTE: All illustrations and photos
have been removed from this article.
More articles from American Journal of Community Psychology
The relationship between perceptions of neighborhood characteristics a..., December 01, 2004
Delinquency and substance use among gang-involved youth: the moderatin..., December 01, 2004
An empowerment evaluation model for sexual assault programs: empirical..., December 01, 2004
Looking for additional articles?
Search our database of over 3 million articles.
Looking for more in-depth information on this industry?
Search our complete database of Industry & Market reports by text, subject, publication
name or publication date.
About Goliath
Whether you're looking for sales prospects, competitive information, company
analysis or best practices in managing your organization,
Goliath can help you meet your business needs.
Our extensive business information databases empower business
professionals with both the breadth and depth of credible,
authoritative information they need to support their business
goals. Whether it be strategic planning, sales prospecting,
company research or defining management best practices -
Goliath is your leading source for accurate information.
|
