Diversity with respect to ethnicity and gender has been studied in open-source and industrial settings for software development. Publication avenues such as academic conferences and journals contribute to the growing technology industry. However, there have been very few diversity-related studies conducted in the context of academia. In this paper, we study the ethnic, gender, and geographical diversity of the authors published in Software Engineering conferences and journals. We provide a systematic quantitative analysis of the diversity of publications and organizing and program committees of three top conferences and two top journals in Software Engineering, which indicates the existence of bias and entry barriers towards authors and committee members belonging to certain ethnicities, gender, and/or geographical locations in Software Engineering conferences and journal publications. For our study, we analyse publication (accepted authors) and committee data (Program and Organizing committee/ Journal Editorial Board) from the conferences ICSE, FSE, and ASE and the journals IEEE TSE and ACM TOSEM from 2010 to 2022. The analysis of the data shows that across participants and committee members, there are some communities that are consistently significantly lower in representation, for example, publications from countries in Africa, South America, and Oceania. However, a correlation study between the diversity of the committees and the participants did not yield any conclusive evidence. Furthermore, there is no conclusive evidence that papers with White authors or male authors were more likely to be cited. Finally, we see an improvement in the ethnic diversity of the authors over the years 2010-2022 but not in gender or geographical diversity.

翻译：族群与性别多样性在开源及工业软件开发环境中已有相关研究。学术会议和期刊等出版渠道对蓬勃发展的科技产业具有推动作用，然而，目前针对学术出版领域多样性问题的研究仍十分匮乏。本文系统研究了软件工程会议与期刊中作者群体的族群、性别及地理分布多样性特征。通过对三本顶级会议（ICSE、FSE、ASE）和两本顶级期刊（IEEE TSE、ACM TOSEM）的出版成果及组委会成员构成进行系统量化分析，结果表明软件工程领域的会议与期刊出版中存在针对特定族群、性别及地理区域作者和委员会成员的偏见与准入门槛。本研究分析了2010至2022年间上述会议期刊的出版数据（已录用作者）与委员会数据（程序委员会/组织委员会/期刊编委会）。数据分析显示，在参与者和委员会成员中，部分群体长期处于显著低比例代表状态，例如非洲、南美洲及大洋洲国家的出版物数量持续偏低。然而，委员会多样性程度与参与者多样性之间的相关性研究未获得确定性结论。此外，未发现关于白人作者或男性作者论文更易被引用的确凿证据。最后，我们观察到2010至2022年间作者群体的族群多样性有所改善，但性别与地理分布多样性未见显著提升。