基于因子分析的长春市犯罪分区研究

选取了2008年长春市主城区各警区(共74个)内的8种常见犯罪,并对其进行了基于主成分的因子分析。基于日常活动理论(routine activities theory)的理论框架,通过分析不同主因子得分的空间分异状况,并结合各地段人口集聚、社会经济、土地利用等方面的特征,探究各主因子的内涵、关联和空间格局的成因。结果表明长春市2008年8种常见犯罪可以归为3个主因子上。其中主因子1在伤害、诈骗和偶发性暴力犯罪上载荷最大,可以归结为经济活动主导(或就业主导型)型因子;主因子2载荷最大的犯罪包括抢劫、抢夺和扒窃,可以归结为商贸、餐饮、文娱活动主导型因子;主因子3载荷最大的犯罪包括入室盗窃和机动车盗窃,可以归结为居住活动主导型因子。

RESEARCH ON CLASSIFICATION OF CRIMES IN CHANGCHUN USING FACTOR ANALYSIS

This study chose Changchun city as the study area to classify the crime areas using the factor analysis based on principal component method, which is conducted in the software of SPSS 16.0. The crime data were collected at the level of police precincts (the number is 74 in total) in Changchun, which consist of Occasional Violent Crimes (including arson, homicide, rape and kidnap), intentional assault, robbery, burglary,auto theft, forcible seizure, pick-pocketing, and fraud. 3 underlying factors are identified as the principal factors which account for more than 70% of the total variance. In order to improve the interpretability of the factors, the solutions are rotated using the varimax orthogonal rotation. The scores of each principal factor is computed and visualized in the maps made by the Arc GIS 9.3, showing the general pictures of criminal situation in Changchun in 2008. The interpretation and connotation of each factor and the reason why the spatial structure and pattern of the principal factors formed are made based on the demographic, socio-economic and land use characteristics for each factors as well as the features of the crimes they have the highest loadings on respectively. The results indicated that the first factor explain 26.99% of the variance, with the largest loadings on intentional assault, fraud, and Occasional Violent Crimes. The second factor account for 25.31% of the variance and has the highest loadings on robbery, forcible seizure and pick-pocketing. The third factor explains 17.93% of the variance with the largest loadings on burglary and the auto theft.