Abstract The use of blacklists, as those that are employed in academic publishing, is problematic with or without biases, but with biases it is even more troubling. This is a theoretical paper that provides insights into the problems of using blacklists for research or for assessing a scholar's output. Biases may appear before an investigation is undertaken as well as after. We also explore the case of a low likelihood of including a non-predatory journal in a blacklist and with a decent power of criteria to detect a true predatory journal. In such an “ideal” list, unfortunately the false discovery rate (FDR) increases as the number of qualities postulated pre-study increases. In addition, the FDR will be biased downwards if the assessor deliberately starts with the assumption that the choice is binary (predatory or not) by lumping qualities together. When bias is introduced also at the post-study level, this further increases the chances that many of the predatory discoveries are false. The issue of bias in predatory publishing has not yet been discussed, and the presence of this factor in the debate on this important issue in academic publishing is relevant to all academics today.

中文翻译：

---

为什么黑名单不可靠：一个理论框架

摘要 学术出版中使用的黑名单的使用，无论有无偏见都是有问题的，但有偏见就更麻烦了。这是一篇理论论文，提供了对使用黑名单进行研究或评估学者产出的问题的见解。偏见可能出现在调查之前和之后。我们还探讨了将非掠夺性期刊列入黑名单的可能性很小的情况，并且有足够的标准来检测真正的掠夺性期刊。在这样的“理想”列表中，不幸的是，错误发现率 (FDR) 会随着假设的预研究质量数量的增加而增加。此外，如果评估者通过将质量混为一谈，故意假设选择是二元的（掠夺性或非掠夺性），则 FDR 将向下偏倚。当在研究后水平也引入偏见时，这进一步增加了许多掠夺性发现是错误的机会。掠夺性出版中的偏见问题尚未讨论，在学术出版这一重要问题的辩论中，这一因素的存在与今天的所有学者都息息相关。