This study employed latent class analysis (LCA) to identify distinct subgroups of U.S. police agencies, each having a similar pattern of criminal investigation goals. Using a nationally representative sample of 1746 U.S. police agencies, including municipal, county sheriffs, and State police, this analysis yielded four subgroups: professional (32.5% of the sample), enforcing (25.2%), servicing (20.5%), and avoiding (21.8%). Results indicate that organizational level predictors of police policies and practices in the criminal investigation process, such as the roles of the public, investigative activities of patrol officers, management of detective work, and forensic support, vary across subgroups. All four variables were significant in predicting membership in the professional agencies when compared to the avoiding agencies. Also, the odds of belonging to the enforcing agencies, compared to avoiding agencies, were significantly increased for agencies with management of detective work and forensic support variables. Furthermore, the servicing agency class membership was significantly related to only forensic support variables, compared to the avoiding agencies. Finally, the mean clearance rates for robberies varied across classes, and the rate was highest in the professional agencies. Police agencies have their own crime-solving methods that predominantly reflect the organizational philosophy and values and also lead to differences in crime clearance rates.

这项研究采用了潜在类别分析（LCA）来识别美国警察机构的不同子组，每个子组都有类似的刑事调查目标模式。使用包括1746个美国警察机构的全国代表性样本，包括市，县警长和州警察，该分析得出四个子类别：专业人员（样本的32.5％），执法（25.2％），服务（20.5％）和避免（21.8％）。结果表明，刑事调查过程中警察政策和实践的组织层面预测因素，例如公众的角色，巡逻人员的调查活动，侦探工作的管理以及法医支持，在各个子组之间有所不同。与回避机构相比，所有四个变量在预测专业机构的成员资格方面均具有重要意义。此外，与管理侦探工作和法医支持变量的机构相比，与避让机构相比，属于执行机构的几率大大增加。此外，服务机构类会员被显著仅涉及法医支持变量，相比避开机构。最后，抢劫的平均清除率在各个类别中各不相同，并且在专业机构中最高。警察机构拥有自己的解决犯罪的方法，这些方法主要反映了组织理念和价值观，也导致犯罪清除率存在差异。