This study proposes a bi-dimensional typology for understanding data-based policy according to whether evidential data are legitimate (legitimacy of data input) and whether the data analytics method used is legitimate (legitimacy of data throughput). The typology is applied to categorize recent cases in Korea and label possible scenarios as one of four types: scientific administration, which analyzes adequate data with an appropriate method (doing the right thing right); data failure, which analyzes inadequate data with an appropriate method (doing the wrong thing right); method failure, which analyzes adequate data with an inappropriate method (doing the right thing wrong); and total fiasco, which uses inadequate data with an inappropriate method (doing the wrong thing wrong). This paper suggests recommendations for making scientific administration cases sustainable, correcting data failures and method failures, and avoiding total fiascos.

这项研究提出了一种二维类型学，用于根据证据数据是否合法（数据输入的合法性）以及所使用的数据分析方法是否合法（数据吞吐量的合法性）来理解基于数据的策略。该类型学用于对韩国最近发生的案件进行分类，并将可能的情况标记为以下四种类型之一：科学管理，以适当的方法分析适当的数据（正确地做正确的事）；数据故障，它使用适当的方法分析不足的数据（正确地做错了事）；方法失败，即使用不适当的方法分析足够的数据（做错正确的事情）；彻底的惨败，它使用不适当的方法使用了不足的数据（做错了错误的事情）。