Abstract:
Theoretical arguments suggest that crime escalates in disadvantaged and disorderly areas because these areas contain cues of danger and safety that signal individuals to stay away, thus reducing effective guardianship, a powerful protective factor against crime. Yet, there is very little knowledge on how perceptions of crime places translate into avoidance or withdrawal behaviors. Moreover, there is limited knowledge of how individual characteristics inform and influence these perceptions. The purpose of this study is twofold. First, this study seeks to understand the accuracy with which people can identify crime hot spots and cool spots within their community. Second, this study will examine the influence of individual predictors on respondents' abilities to identify crime and non-crime locations within the two study neighborhoods. Specifically, individual level predictors of individual demographics, perceptions of crime and disorder, and neighborhood familiarity and tenure are tested. Study measures are derived using two data sources including cognitive maps administered to active community members
(N=168) through the Communities Problems and Issues Survey (CPIS) and calls for service to the Trinidad and Tobago Emergency Response System (E-999). Accuracy and the influence of individual predictors are tested using a mix of analytic techniques including descriptive diagnostics, t-tests, zero-inflated count regression analysis and ordinal logistic regression. Overall, the study supports past perception of crime research by determining that respondents are not accurate in identifying crime hot spots. Additionally, when testing the individual predictors that influence accuracy, two factors, gender and neighborhood familiarity, have a strong influence on whether respondents include crime hot spots in the areas they consider unsafe or dangerous. The study concludes with a discussion of the study's implications for both practice and research.
