Inmate murder is a grave threat to institutional safety in correctional settings, unfortunately relatively little prior research has studied it. The current study analyzed data from 636 capital murderers sentenced to death in California of whom 6% had murdered other inmates during their confinement career. Bivariate analyses found that inmate murderers had more extensive and violent offending histories, greater security threat group involvement, more institutional misconduct, were disproportionately white, and exhibited greater and more diverse psychopathic features relative to inmates who did not murder. Logistic regression model found that interpersonal and affective psychopathic features, security threat group, white race, and institutional misconduct history were significantly associated with prison murder. Prior murder convictions, psychopathy total score, security threat group activity, institutional misconduct, and a multiplicative term for security threat group members with psychopathy had adequate to excellent classification accuracy in a ROC-AUC model. We encourage similar data collection efforts with condemned populations to specify risk factors for individuals most likely to perpetrate murder while in prison custody.

囚犯谋杀是对惩教环境中机构安全的严重威胁，不幸的是，之前的研究相对较少。目前的研究分析了加利福尼亚州 636 名被判处死刑的死刑犯的数据，其中 6% 的人在监禁期间曾谋杀过其他囚犯。双变量分析发现，与没有谋杀的囚犯相比，囚犯杀人犯有更广泛和暴力的犯罪历史、更多的安全威胁团体参与、更多的机构不当行为、不成比例的白人，并且表现出更多和更多样化的精神病特征。逻辑回归模型发现，人际和情感精神病特征、安全威胁群体、白人种族和机构不当行为史与监狱谋杀案显着相关。之前有过谋杀定罪，精神病总分、安全威胁团体活动、机构不当行为以及患有精神病的安全威胁团体成员的乘法术语在 ROC-AUC 模型中具有足够出色的分类准确性。我们鼓励对被判死刑的人群进行类似的数据收集工作，以确定在监狱拘留期间最有可能实施谋杀的个人的风险因素。