In designing risk assessment algorithms, many scholars promote a "kitchen sink" approach, reasoning that more information yields more accurate predictions. We show, however, that this rationale often fails when algorithms are trained to predict a proxy of the true outcome, as is typically the case. With such "label bias", one should exclude a feature if its correlation with the proxy and its correlation with the true outcome have opposite signs, conditional on the other model features. This criterion is often satisfied when a feature is weakly correlated with the true outcome, and, additionally, that feature and the true outcome are both direct causes of the remaining features. For example, due to patterns of police deployment, criminal behavior and geography may be weakly correlated and direct causes of one's criminal record, suggesting one should exclude geography in criminal risk assessments trained to predict arrest as a proxy for behavior.

翻译：在设计风险评估算法时，许多学者推崇"万全之策"的方法，其理由是：更多信息能带来更准确的预测。然而，我们证明，当算法被训练用于预测真实结果的代理指标（这是典型情况）时，这一逻辑往往失效。面对此类"标签偏差"，应排除某一特征的条件是：在控制模型其他特征后，该特征与代理指标的相关性及其与真实结果的相关性符号相反。当特征与真实结果弱相关，且该特征与真实结果共同成为剩余特征的直接原因时，该条件往往成立。例如，由于警力部署模式，犯罪行为与地理位置可能弱相关，并共同成为犯罪记录的直接原因，这表明在训练以逮捕作为行为代理进行预测的刑事风险评估中，应排除地理位置这一特征。