National diversity at conferences of the International Communication Association

ABSTRACTAcademic conferences, particularly international conference venues, have been suggested as a way to overcome inequality within academic associations. In this study, we assess the national diversity of 18 annual conferences of the International Communication Association (ICA) between 2005 and 2022 and compare the composition of conference contributors to that of ICA members. Our results show that although the overall national diversity of ICA conferences has more than doubled since 2005, and closely mirrors ICA membership, the effective number of contributing countries is still very low. National diversity differs substantially between divisions and depending on the conference venue, with non-US venues exhibiting significantly more national diversity. We discuss measures to evaluate and improve ICA’s national diversity at different levels.KEYWORDS: DiversityrepresentationconferenceInternational Communication Associationscientometrics Disclosure statementNo potential conflict of interest was reported by the author(s).Data availability statementOpen Science. Replication data and code can be foundhere: https://osf.io/3pe6y/.Notes1 We chose the year 2005 as a starting point for our analysis, since the conference programs were accessible online and we were able to pull the full population of conference contributors and contributions.2 The data set is part of a larger project and was also used for a study focusing on the gender diversity of ICA conferences (Braun, Heintz, Kruschinski, Trepte & Scharkow, Citationaccepted for publication).3 We thank ICA headquarters for their support in provision of data on conference contributions from 2019 to 2022 and anonymized membership data for 2020.4 As several reviewers suggested testing a diversity measure beyond the country level, we also analyzed our data using Demeter’s (Citation2018) classification of countries into six regions, for which we then computed diversity scores. These regional diversity scores were somewhat lower, but also less dispersed: Across all years and divisions (n = 515), the mean country diversity was 2.96 (SD = 1.48) and the mean regional diversity was 2.42 (SD = 0.80). The correlation between both measures was r = .91, and the model estimates were virtually identical. We provide the results for this alternative measure in the supplementary materials.5 Without exclusion of single-authored presentations, differences between divisions in presentation-level diversity would strongly depend on the relative frequency of single-authored presentations, which varied from 86% in the Communication History division to 12% in the Communication Science and Biology division.6 For this analysis, we counted thematic and sponsored sessions as well as interest groups as conference divisions, which resulted in k = 35 different levels for this grouping variable.7 Since the ICA conference in 2020 was scheduled to take place in Australia, we expect that a larger than usual number of Australian scholars became members for the purpose of attending.