Physical environment and crime and misconduct in Kentucky schools.

Drawing upon ecological theories of crime control, aspects of the physical environment such as building design, street layout, and land use are thought to indicate territoriality and natural surveillance, thereby affecting the ability of residents to supervise and intervene in crime. To date, ecological models have been tested primarily at community levels of analysis (i.e., neighborhood, block). In contrast, this paper tests the applicability of this theoretical approach to crime in school settings. More specifically, we estimate random-intercept models using survey data from 3682 7th-grade students and 1351 teachers, nested within 65 Kentucky schools linked to school-level measures of the physical environment to determine how they affect various measures of school-based crime and misconduct. Editors' Strategic Implications:How one measures school violence may have profound effects on what contributing causal factors are identified and strategies for prevention are developed. Student reports of school violence appear linked to more normative daily behaviors, whereas teacher reports-though limited to officially observed behaviors-are typically more serious. Thus, measurement implications abound. Nevertheless, territoriality, surveillance, and a sense of order, particularly in the immediate school context more so than the larger community context, appear linked to school violence.