This study explores the characteristics of scientific activity patterns through co-author affiliations to obtain new insights into interdisciplinary research. To classify the interdisciplinarity in research, we explored and compared two different approaches: the diversity of disciplines reflected in the listed affiliations of the authors and the diversity of the subject categories reflected in the reference list. To assess the diversity in departmental affiliations, we developed an explorative methodology that retrieves feature words from a combination of manual work and the thesaurus function in the Thomson Data Analyzer text mining tool. To assess the diversity in references, we followed the conventional approach applied in previous work. With both approaches, we relied on diversity as the measure for assessing interdisciplinarity of 157,710 articles published in PloS One (2007–2016). Based on a comparison between the results of both approaches, our study confirms that different methodologies and indicators can produce seriously inconsistent, and even contradictory results. In addition, different indicators may capture different understandings of such a multi-faceted concept as interdisciplinarity. Our results are summarized in a schematic representation of this twofold perspective as a method of indexing the different types of interdisciplinarity commonly found in research studies.

The present study is an extended version of an article presented at the 16th International Conference on Scientometrics and Informetrics, Wuhan (China), 16–20 October 2017 (Zhang et al. 2017). The authors would like to acknowledge support from the National Natural Science Foundation of China (Grant No. 71573085), the Innovation Talents of Science and Technology in HeNan Province (Grant Nos. 16HASTIT038, 2015GGJS-108), and the Excellent Scholarship in Social Science in HeNan Province (No. 2018-YXXZ-10). We thank Giovanni Abramo and Ronald Rousseau for inspiring discussions.

"This is a post-peer-review, pre-copyedit version of an article published in Scientometrics. The final authenticated version is available online at: https://doi.org/10.1007/s11192-018-2853-0”.

Peer reviewed