Cognitive models of psychosis have stimulated empirical studies on cognitive biases involved in schizophrenia spectrum psychoses and their symptoms. This systematic review aimed to summarize the studies on the role of cognitive biases as assessed in different performance-based tasks in schizophrenia spectrum psychoses and clinical high-risk states. We focused on five cognitive biases linked to psychosis, i.e., aberrant salience, attentional biases, source monitoring biases, jumping to conclusions, and bias against disconfirmatory evidence. We identified = 324 studies published in = 308 articles fulfilling inclusion criteria. Most studies have been cross-sectional and confirmed that the schizophrenia spectrum psychoses are related to exaggerated cognitive biases compared to healthy controls. On the contrary, less evidence suggests a higher tendency for cognitive biases in the UHR sample. The only exceptions were source monitoring and jumping to conclusions, which were confirmed to be exaggerated in both clinical groups. Hallucinations and delusions were the most frequent symptoms studied in the context of cognitive biases. Based on the findings, we presented a hypothetical model on the role of interactions between cognitive biases or additive effects of biases in shaping the risk of psychosis. Future research is warranted for further development of cognitive models for psychosis.

中文翻译：

---

对精神分裂症谱系精神病和临床高风险状态认知偏差的基于绩效的评估研究的系统回顾：40年研究总结

精神病的认知模型激发了对精神分裂症谱系精神病及其症状所涉及的认知偏差的实证研究。本系统综述旨在总结在精神分裂症谱系精神病和临床高风险状态的不同基于表现的任务中评估认知偏差的作用的研究。我们重点关注与精神病相关的五种认知偏差，即异常显着性、注意力偏差、源监控偏差、仓促下结论以及对否定性证据的偏差。我们确定了 = 324 项研究发表在 = 308 篇满足纳入标准的文章中。大多数研究都是横断面的，并证实与健康对照相比，精神分裂症谱系精神病与夸大的认知偏差有关。相反，较少的证据表明 UHR 样本中认知偏差的倾向较高。唯一的例外是源头监测和仓促下结论，这在两个临床组中都被证实被夸大了。幻觉和妄想是在认知偏差的背景下研究的最常见的症状。根据这些发现，我们提出了一个假设模型，探讨认知偏见之间的相互作用或偏见的附加效应在塑造精神病风险中的作用。未来的研究有必要进一步开发精神病的认知模型。