Spatial Configuration and Actual Crime Locations in a University Campus Setting Yixiang Long and Perver K. Baran College of Design, NC State University Paper presented at EDRA37, Atlanta, 3-7 May, 2006
Overview Introduction Study Objectives Methodology Results Conclusions
Introduction Crime has long been thought to be intimately associated with the physical environment. A body of research in this area has focused on exploring the relationship between physical environment and fear of crime (Wilson & Kelling, 1982); Lewis & Salem, 1985; Perkins et al., 1992). Another body of research has focused on exploring the relationship between physical environment and actual crime locations (Beavon et al., 1994; Loukoaitou-Sideris, 1999). More recently, research has examined the relationship between the occurrence of criminal events and spatial configuration by using Space Syntax methodology (Hiller, 1998; Hillier and Sahbaz, 2005; Shu, 1999, Nubani & Wineman, 2005).
Problem and purpose of the study The controversial results point to the fact that the relationship between space and crime occurrence is not a simple one. In addition, so far most of the space syntax research has focused on residential neighborhoods or housing estates. Purpose: Investigate the relationship between spatial configuration and actual crimes in a campus environment, NC State University campus.
Study Area Main campus of NC State University (Raleigh, NC)
Hypotheses Outdoor crimes occur in areas close to the campus buildings and roads, since there are more opportunities for crime (potential victims). It is assumed that outdoor crimes cluster in more integrated and connected areas. There is positive correlation between integration values, connectivity, control values and crime occurrences along the axial lines. Based on the theory of prospect-refuge, the hot spots of crimes in campus are areas that have more potential of concealment, blocked prospect and blocked escape.
Methodology Stage 1 --- Campus Wide Analysis Crime incident data from 2003-2004 Configurational analysis - Space Syntax Identifying hot spots and cold spots - CrimeStat Stage 2 --- Site analysis Qualitative evaluation of environmental features in hot spots and cold spots Visibility analysis - DepthMap
Results Stage 1 Frequency of parking deck and outdoor crime events Total outdoor crime events: 495 (parking decks and outdoor environment) Parking decks: 48 crime events (10%) Outdoor environment: 447 crime events (90%)
Frequency of outdoor crime types
The relationship between outdoor crimes, campus buildings and roads 45% Distance to building Distance to road Outdoor crimes occur closely to the buildings and roads, where there is more movement and more potential victims. In addition, roads could be used as access and escape routes for offenders.
Configurational Analysis Global Integration Local Integration Control value Connectivity
The relationship between global integration and outdoor crimes Unit of analysis: axial line
Bivariate correlations between crime count (dependent) and Space Syntax parameters (independent variables) Crime Connectivity Control Count Crime Count 1 Connectivity 0.48 1 Control 0.6 0.86 1 Local Integration Global Integration Local Integration 0.27 0.68 0.4 1 Global Integration 0.33 0.35 0.21 0.72 1 Axial Line Length (ft) 0.72 0.69 0.77 0.4 0.37 1 N = 285, All correlations are significant at the 0.01 level (2-tailed). Axial Line Length (ft) Further multiple regression model needed to determine the extent that that differences in the syntactical properties might contribute to differences in the crime count.
Regression on Crime Count Standardized Regression Coefficients (Beta) Global Integration Local Integration Axial Line Length Control.220*** -.194**.561***.195** R 2 = 0.539, P =.000, N = 285 Larger axial line length more crime Larger global integration more crime Larger degree of control more crime Larger local integration less crime
Results Stage 2 The relationship between hot spots and environmental features Hot spot analysis CrimeStat
Hot spot analysis results CrimeStat Hot spots in Campus (measured by CrimeStat) Area Crime Density (frequency/ sqft) *10,000 Hot Spot Wood 926 Student center 355 Medium Bowen 337 Bragaw 296 Cold Spot POE 0 Arboretum 0
Hot Spots, Medium Spots and Cold Spots
Qualitative Evaluation of Hot spots and Cold Spots According to Fisher and Nasar (1992), the concepts of prospect-refuge theory could be describes in relation to crime as the following: 1. Concealment or refuge: criminal can hide, not easily to be seen by potential victims or others. 2. Blocked prospect: places with expansive view without been seen. 3. Blocked escape: features that block exits or connections to others who might intervene in case of attack This theory is used to qualitatively evaluate the environmental features in hot spots and cold spots.
Hot spots Medium Spots Cold spots Wood and Student Center Bowen and Bragaw Poe and Arboretum
Hot spots Student Center Wood Student center: complex topography; limited connection to the other places; combination of retailing wall, vegetation and ramps offers various refuge places; more integrated area with more movements; lack of natural surveillance. Wood: More segregated area with low integration values; woods and trees blocking escape routes for victims and views outside.
Medium Bowen Bragaw More open space with expansive view, as residential hall, lots of potential victims walking around the area, combination of small windows and window blinds make these places low in natural surveillance.
Cold spots Poe Arboretum More connections, more natural surveillance, close to large parking lots and on-street parking
Visibility Analysis Controllability Visual control: Visually dominant area, easily see people s activities without difficulty being seen by observed people. Visual controllability: Visually dominated area, people are not aware that the others observe their activities (Turner, 2004).
Wood Poe Visual control: no differences between hot spots, medium and cold spots. Visual controllability Hot spot (Wood) : 0.66 Medium (Bragaw): 0.56 Cold spot (Poe): 0.3 Bragaw
Conclusions Crime in a bounded open space, such as a campus, has different dynamics than in residential and neighborhood environments where sociodemographic and land use variables are important. Offenders in a campus setting look for opportunities, i.e. large number of potential victims, such as areas close to buildings and streets with high integration values. However, offenders also look for areas that provide opportunities for escape, i.e. high control and connectivity values. Combination of the physical features can potentially provide more refuge places with blocked escape routes as well as blocked view, i.e. neighboring routes with less connectivity and high visual controllability. Although the findings are not conclusive, this study has shown that visibility analysis has potential to analyze site specific dynamics of crime occurrence.
Thank you!