A LONGITUDINAL ANALYSIS OF NEIGHBORHOOD CRIME RATES USING LATENT GROWTH CURVE MODELING
A LONGITUDINAL ANALYSIS OF NEIGHBORHOOD CRIME RATES USING LATENT GROWTH CURVE MODELING
复制标题
DOI：
10.1525/sop.2010.53.1.127
复制
发表时间：
2010-03-01
影响因子：
2.4
通讯作者：
Desmond, Scott A.
中科院分区：
文献类型：
--
作者：
Kikuchi, George;Desmond, Scott A.
虽然社会解体理论为理解社区变化如何影响犯罪率提供了一个框架，但对社区特征和犯罪的研究主要依赖于横截面数据。利用潜在增长曲线模型和印第安纳波利斯住宅盗窃和车辆盗窃的纵向数据，作者分析了社区特征与犯罪趋势之间的关系，这些数据在1992年至2006年期间每年在人口普查街区水平上进行测量。对于住宅盗窃和车辆盗窃，基线模型显示，犯罪率的变化最好用一个二次函数来描述，即最初的线性下降和随后的减速（减慢）下降。当社区特征作为预测因素时，社区劣势的变化与住宅入室盗窃和车辆盗窃的变化随时间的变化显著相关，而住宅稳定性的变化对住宅入室盗窃和车辆盗窃的变化没有显著影响。
Although social disorganization theory provides a framework for understanding how changes in neighborhoods can influence crime rates over time, research on neighborhood characteristics and crime has relied primarily on cross-sectional data. Using a latent growth curve model and longitudinal data on residential burglary and vehicle theft in Indianapolis, measured annually between 1992 and 2006 at the census block group level, the authors analyzed the relationships between neighborhood characteristics and crime trends. For both residential burglary and vehicle theft, baseline models revealed that changes in crime rates were best captured by a quadratic function with an initial linear decrease and subsequent deceleration ( slowing) of the decrease. When neighborhood characteristics were included as predictors, change in neighborhood disadvantage was significantly related to changes in both residential burglary and vehicle theft over time, while change in residential stability did not have a significant effect on changes in residential burglary or vehicle theft.