Dangerousness Index for Ecuadorian Persons Deprived of Liberty: A Multidimensional Social Indicator Using CATPCA and K-Means

This dataset contains the data and supporting documentation used in the study “Dangerousness Index for Ecuadorian Persons Deprived of Liberty: A Multidimensional Social Indicator Using CATPCA and K-Means.” The research is based on de-identified secondary data from the 2022 Ecuadorian Penitentiary Census, conducted by the National Institute of Statistics and Censuses of Ecuador (INEC). The repository includes the original working database and codebook, the selected and transformed analytical variables, variable definitions and coding schemes, the statistical procedures used to perform categorical principal component analysis (CATPCA), index normalization and K-means classification, as well as the principal analytical outputs. It also contains the official census questionnaire and methodological documentation required to understand the source, structure and processing of the data. The final analytical dataset comprises 31,273 valid records and 52 selected variables covering sociodemographic characteristics, legal and penal trajectory, previous institutional contact, prison living conditions, access to services, programme participation and social-support factors. CATPCA was used to quantify categorical variables and construct a normalized institutional risk indicator ranging from 0 to 1. K-means clustering was subsequently applied to classify the observations into five ordered levels: very low, low, moderate, high and very high. These materials are provided to facilitate transparency, verification and reproducibility of the data-processing and statistical-analysis procedures. The resulting indicator is intended as a descriptive and institutional-management support measure. It should not be interpreted as a deterministic prediction of future violence, recidivism or individual dangerousness, nor should it replace professional and individualized assessment.本数据集包含研究《厄瓜多尔被剥夺自由者危险性指数：基于CATPCA与K均值聚类的多维社会指标》所用的数据与辅助文档。本研究基于厄瓜多尔国家统计与普查研究所（INEC）开展的2022年厄瓜多尔监狱人口普查的去标识化二手数据。 本存储库包含原始工作数据库与代码手册、筛选并转换后的分析变量、变量定义与编码方案、用于开展分类主成分分析（CATPCA）、指数标准化及K均值分类的统计流程，以及主要分析结果。此外还包含官方普查问卷与帮助理解数据来源、结构及处理流程的方法学文档。 最终分析数据集包含31273条有效记录与52个筛选变量，覆盖社会人口学特征、法律与刑罚轨迹、既往机构接触史、监狱生活条件、服务获取情况、项目参与情况及社会支持因素。研究采用CATPCA对分类变量进行量化，并构建取值范围为0至1的标准化机构风险指数。随后通过K均值聚类将观测样本划分为五个有序分级：极低、低、中等、高及极高。 本材料旨在提升数据处理与统计分析流程的透明度、可验证性与可复现性。所得指数仅作为描述性及机构管理支持工具，不应被解读为对未来暴力行为、再犯或个体危险性的确定性预测，亦不得替代专业化个性化评估。