Abstract The dark figures of crime are occurrences that, by some criteria, are called crime yet that are not registered in the official statistics. According to several studies, only a few rapes are reported to the authorities. The current study, using crime data from Campinas, Brazil, sought to examine the spatial dark figures of rapes through the comparison between the spatial patterns of incidents from an official source (police) and the spatial patterns of incidents from an unofficial source (hospital). We used Kernel density estimation maps, generalized Gini coefficient, and a spatial point patterns test to measure the spatial dissimilarities between both sources. Also, we estimated the likelihood of spatial dark figures of rapes using logistic regression models. The results indicate patterns of spatial dark figures of rapes and its association with the street segment and the neighborhood factors. The findings suggest the potential for partnerships between police and medical services in targeting locations with high levels of rape underreport. In addition, it supports place-based prevention measures.

中文翻译：

---

强奸的空间暗数字：（In）警察和医院数据的一致性

摘要 犯罪的黑暗数字是指根据某些标准被称为犯罪但未在官方统计中登记的事件。根据几项研究，只有少数强奸案被报告给当局。目前的研究使用来自巴西坎皮纳斯的犯罪数据，试图通过比较官方来源（警察）的事件空间模式和非官方来源（医院）的事件空间模式来检查强奸的空间暗图. 我们使用核密度估计图、广义基尼系数和空间点模式测试来测量两个来源之间的空间差异。此外，我们使用逻辑回归模型估计了强奸的空间暗图的可能性。结果表明强奸的空间暗图模式及其与街道段和邻里因素的关联。调查结果表明，警察和医疗服务部门之间有可能建立伙伴关系，以针对强奸率低报率高的地区。此外，它还支持基于地点的预防措施。