Computational Communication Science: A Methodological Catalyzer for a Maturing Discipline
Martin Hilbert, George A. Barnett, Joshua Evan Blumenstock, Noshir Contractor, Jana Diesner, Seth D. Frey, Sandra González‐Bailón, PJ Lamberso, Jennifer Pan, Tai‐Quan Peng, Cuihua Shen, Paul E. Smaldino, Wouter van Atteveldt, Annie Waldherr, Jingwen Zhang, Jonathan Jh Zhu · CityU Scholars · 2019
This article reviews the opportunities and challenges for computational research methods in the field of communication.Among the social sciences, communication stands out as a discipline with a relatively low-profile institutionalized focus on the in-house development of methods.Computational tools are changing this, and they are catalyzing a new set of methods directly suited to tackling foundational research questions in communication.We systematically review how computational methods affect the three fundamental pillars of the scientific method: observational approaches (i.e., digital trace data), theoretical 1 The conversation culminating in this article started at the conference "Re-Computing Social Sciences: What Have We Learned After a Decade?" held on May 19, 2017, at the University of California, Davis, with the assistance of the UC Davis Institute for Social Sciences, the UC Davis Department of Communication, and the UC Davis Computational Communication Research Lab.It benefited from the discussions and presentations of the panels organized by the computational methods interest group at the 2017 and 2018 annual conferences of the International Communication Association in San Diego and Prague.We would also like to thank several blind peer reviewers and the editors of this Special Section, especially JungHwan Yang, who would certainly deserve to be accredited as co-authors in this collective effort, given their detailed and fruitful contributions to this cooperative work.Martin Hilbert et al.International Journal of Communication 13( 2019) 3914 approaches (i.e., computer simulations), and experimental research (i.e., virtual labs and field experiments).We stress that data are a catalyzer but not a requirement for computational science.We explore how observational, theoretical, and experimental approaches can be combined and cross-fertilize one another.We conclude that taking advantage of computational methods will require a systematic effort in our discipline to develop and adjust these methods.