The effect of comment moderation on perceived bias in science news

ABSTRACT

Uncivil comments following online news articles about issues of science and technology have been shown to lead to biased interpretations of the news content itself. Using an experiment embedded in a nationally representative survey, we provide evidence that cues about comment moderation ‒ even without any change in the comments themselves ‒ have the potential to alleviate this so-called nasty effect. Participants exposed to uncivil comments that appear in a moderated environment were less likely to perceive bias in the news article itself. Importantly, perceptions of bias among respondents exposed to the uncivil, moderated stimulus were comparable to those of respondents who viewed both moderated and unmoderated civil comments. Our results suggest that visible cues about comment moderation are a potentially valuable endeavor for news organizations, especially in an age of declining profit margins.

KEYWORDS:

• Nasty effect
• incivility
• online comments
• comment moderation
• perceived bias

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes on contributors

Sara K. Yeo (Ph.D., University of Wisconsin-Madison) is an Assistant Professor in the Department of Communication at the University of Utah. Her research interests include science communication, public opinion of STEM issues, and information seeking and processing. [email: [email protected]]

Leona Yi-Fan Su (Ph.D., University of Wisconsin-Madison) is an Assistant Professor in the Department of Communication at the University of Utah. Her research interests focus on the interplay between new media and society, particularly in the context of science and environmental communication. [email: [email protected]]

Dietram A. Scheufele (Ph.D., University of Wisconsin-Madison) is the John E. Ross Professor in Science Communication and Vilas Distinguished Achievement Professor at the University of Wisconsin-Madison and in the Morgridge Institute for Research. His research deals with the public and political interfaces of emerging science. [email: [email protected]]

Dominique Brossard (Ph.D., Cornell University) is Professor and Chair in the Department of Life Science Communication at the University of Wisconsin-Madison. Her research agenda focuses on the intersection between science, media and policy. [email: [email protected]]

Michael A. Xenos (Ph.D., University of Washington) is Communication Arts Partners Professor and Chair in the Department of Communication Arts at the University of Wisconsin-Madison. His primary research focus is on the extent to which the internet and social media may help individuals learn about political issues, form opinions, and participate in politics. [email: [email protected]]

Elizabeth A. Corley (Ph.D., Georgia Institute of Technology) is Professor in the School of Public Affairs at Arizona State University. Her research interests focus on technology policy and environmental policy. Her book titled ‘Urban Environmental Policy Analysis’ (with Heather E. Campbell) was published by M.E. Sharpe in 2012. [email: [email protected]]

ORCID

Sara K. Yeo http://orcid.org/0000-0002-2043-8400

Notes

1 Controls for the number of comments were not included in the regression model as there should be no systematic differences in our dependent variable due to random assignment. In order to confirm this, we ran the model controlling for number of comments. There were no substantive differences between the models with and without the control.

2 Additionally, we ran an OLS regression model that included demographic controls, but there were no substantive differences between the models with and without these controls.

3 Significant interactions were further examined using a conditional process modeling program, PROCESS, which is an add-on to SPSS (Hayes, 2013; Hayes & Matthes, 2009). The PROCESS macro cannot accommodate variable weights. Fortunately, PROCESS output is equivalent to that of an OLS regression in SPSS with listwise deletion (Hayes, 2013). Therefore, we ran a regression model using SPSS with the weighted data. The results of both PROCESS and SPSS (weighted) models were similar and included statistically significant interactions between exposure to uncivil comments and moderation. We chose to keep the PROCESS model to probe the interaction but report the results of weighted regression model in Table 1. The unstandardized predicted values of the dependent variable from the weighted model were also used to graph the interaction (Figure 2(A)).

Additional information

Funding

This work was supported by the National Science Foundation [grant numbers SES-0937591 and SES-DMR-0832760]. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.