The Effects of Automated Journalism on Hostile Media Biases

• In Event: Poster Session: Health Policy and Politics & Information Technology and Politics
  In Poster Session: Poster Session: Information Technology and Politics

Abstract

Automated journalism (AJ), is the generation of news by algorithms with little or no human interference (Graefe, 2016). AJ is currently applied mostly to data-heavy fields, but is considered for reporting on political and controversial topics as well. This change could shape the relationship between audiences and media, at a time of low trust in journalism. The goal of this study is to examine whether AJ could ameliorate perceptions of journalistic bias and hostility among readers in the context of gun control.
The perceptions of journalistic bias independently of actual slant is known as the Hostile Media Bias (HMB), where partisans are prone to perceive media coverage as biased against their side, even when the articles they read were judged by non-partisans to be objective and balanced (Perloff, 2015). Since HMB is influenced by articles’ source (Gunther et al., 2016), we argue that the growing application of automated news production could change public perceptions of hostility, as studies found that readers often perceive algorithms to be less biased than humans, a cognitive shortcut known as the “machine heuristic” (Cloudy et al., 2021).
In this study, we expect AJ to produce lower levels of HMB compared to human journalists (H1), and that effect will be mediated by activation of the machine heuristic (H2). Based on prior work by Gunther & Liebhart (2006), we assume both AJ and human journalists yield more HMB than college students (H3). We expect HMB in AJ to be moderated by perceived novelty of the new technology and algorithmic literacy (Dogruel, 2021). However, it is not clear whether algorithmic literacy will increase or decrease HMB. On one hand, those familiar with AJ might perceive it as less novel and therefore less frightening. On the other hand, those high on algorithm literacy are more likely to be familiar with the fact that algorithms are not completely independent from the programmers, and are thus influenced by human biases. We thus pose the moderating effect of algorithm literacy as a research question and not a directional hypothesis (RQ1).
To examine these questions, 150 students from a large public U.S. northeastern university will be recruited and randomly assigned to one of three conditions; college student, human journalist, or algorithm journalist. Prior work compared articles written by AJ to those written by humans in an attempt to secure ecological validity. Such decision prevented the establishment of causal impact of mere source effect, as conditions differed in content and not only source. Therefore, we will expose all participants to the same articles, and manipulate only the stated source of articles. The news articles will be about the politically polarizing topic of gun control, and will contain equal number of pro- and anti- gun control arguments. Articles will be piloted and refined to make sure non-partisans perceive them as neutral and unbiased. This study would contribute to our understanding of changes in HMB in light of recent technological developments in journalism, and could inform practitioners and policymakers wishing to reduce perceptions of bias among partisans and regain trust in media and journalism.
References
Cloudy, J., Banks, J., & Bowman, N. D. (2021). The Str(AI)ght Scoop: Artificial Intelligence Cues Reduce Perceptions of Hostile Media Bias. Digital Journalism, 1–20. https://doi.org/10.1080/21670811.2021.1969974
Dogruel, L. (2021). What is Algorithm Literacy? A Conceptualization and Challenges Regarding its Empirical Measurement. In M. Taddicken & C. Schumann (Eds.), Algorithms and Communication (Vol. 9, pp. 67–93). https://doi.org/10.48541/dcr.v9.3
Graefe, A. (2016). Guide to Automated Journalism. https://doi.org/10.7916/D80G3XDJ
Gunther, A. C., & Liebhart, J. L. (2006). Broad Reach or Biased Source? Decomposing the Hostile Media Effect. Journal of Communication, 56(3), 449–466. https://doi.org/10.1111/j.1460-2466.2006.00295.x
Gunther, A. C., McLaughlin, B., Wise, D., & Wise, D. (2016). Who Says What to Whom: Content Versus Source in the Hostile Media Effect. International Journal of Public Opinion Research, edw009. https://doi.org/10.1093/ijpor/edw009
Jamieson, K. H., & Cappella, J. N. (2008). Echo Chamber: Rush Limbaugh and the Conservative Media Establishment. Oxford University Press.
Perloff, R. M. (2015). A Three-Decade Retrospective on the Hostile Media Effect. Mass Communication and Society, 18(6), 701–729. https://doi.org/10.1080/15205436.2015.1051234

Authors

• Rui Wang, University at Buffalo

• Yotam Ophir, University at Buffalo