Social cognition remains one of the most difficult cognitive domains to assess in diverse populations due to a lack of culturally appropriate tools. This study systematically reviewed literature on neuropsychological tests for social cognition that have been translated, adapted, are cross-cultural, or are assembled for diverse, specifically “Global South,” populations. The aim was to identify assessments appropriate for diverse populations, outline and evaluate their methodological approaches, and provide procedural recommendations for future research. The PRISMA systematic review search strategy produced 10,957 articles, of which 287 were selected for full-text screening. The study had to include a neuropsychological assessment of social cognition. The full text of the resulting 287 articles was then screened; the study had to include a translated, adapted, cross-cultural test, or an assembled test for Global South populations. Eighty-four articles were included in this study: 24 for emotion recognition, 45 for theory of mind, 9 for moral reasoning, and six for social cognition in general. Overall, there were 31 translations, 27 adaptations, 14 cross-cultural tests, and 12 assembled tests for Global South populations. Regarding quality, 35 were of low quality, 27 were of moderate quality, and 22 were high quality. This study provides an overview of social cognition tests modified or assembled for diverse populations and gives examples of methodological procedures. It highlights the variability in procedure quality and provides possible reasons for this variability. Finally, it suggests a need to report rigorous modification and assembly procedure in order to have modified and assembled social cognition tests appropriate for diverse populations.

由于缺乏适合文化的工具，社会认知仍然是在不同人群中评估最困难的认知领域之一。这项研究系统地回顾了有关社会认知的神经心理学测试的文献，这些文献已经被翻译、改编、跨文化或针对不同的人群，特别是“全球南方”人群而汇集。目的是确定适合不同人群的评估，概述和评估其方法，并为未来的研究提供程序建议。PRISMA 系统综述检索策略产生了 10,957 篇文章，其中 287 篇被选择进行全文筛选。该研究必须包括对社会认知的神经心理学评估。随后筛选了 287 篇文章的全文；该研究必须包括针对全球南方人口的翻译、改编、跨文化测试或组合测试。这项研究纳入了 84 篇文章：24 篇关于情感识别，45 篇关于心理理论，9 篇关于道德推理，6 篇关于一般社会认知。总体而言，共有 31 项翻译、27 项改编、14 项跨文化测试以及 12 项针对全球南方人口的组装测试。从质量来看，低质量35个，中等质量27个，优质22个。本研究概述了针对不同人群修改或组装的社会认知测试，并给出了方法程序的示例。它强调了程序质量的可变性，并提供了这种可变性的可能原因。最后，它建议需要报告严格的修改和组装程序，以便修改和组装适合不同人群的社会认知测试。