Individual Submission Summary
Share...

Direct link:

Download

Eye-tracking the News: Comment Disagreement and News Shareability on Facebook

Fri, August 30, 4:00 to 5:30pm, Hilton, Embassy

Abstract

Current research predominantly examines the content of news articles and tests the effect of controversial framing on user engagement via social media. The findings from this literature tend to show that controversial topics and frames increase user commenting on news articles (Ksiazek, 2018; Ziegele et al., 2018), but controversy decreases user-generated news sharing on social media (Valenzuela et al 2017). Agreement -- the opposite of controversy -- increases the probability that news articles are shared on social media. While previous research has convincingly argued that controversy in news articles drives commenting but decreases sharing but may increase reading (Bright 2016), no study has examined the effect of controversy within comments on users’ propensity to share or read news on Facebook.

In this paper, we examine how controversy in Facebook comments influences citizens’ likelihood to read and share news on the platform. Focusing on comments is particularly relevant in the context of social media for two reasons. First, commenting allows citizens’ opinions to appear alongside traditional editorial cues, and therefore comments may serve as a heuristic shortcut in evaluating and selecting news on Facebook (Winter, 2018; Neubaum & Krämer, 2017). Second, the act of commenting signals relevance to Facebook’s algorithm, which influences the visibility of a news item across the platform (Bucher, 2012).

Political news is one of the most controversial news genres, since politics is characteristically about the competition for the power to rule. Therefore, we test the effect of comment controversy on user behavior across five genres of news (politics, economy, society, culture, and sports), but maintain an empirical focus on political news.

We formulate three main hypotheses:
H1: Posts in the disagreeing comments condition will be related to higher levels of attention
H1a: Posts in the disagreeing comments condition will be related to higher levels of attention when post content is political.
H2: Posts in the disagreeing comments condition will be associated with lower likelihood to share post
H2a: The negative effect of disagreement in comments on likelihood to share will be strongest for political content compared to other content genres.
H3: Posts in the disagreeing comments condition will be associated with higher likelihood to read post
H3a: The positive effect of comment disagreement on likelihood to read will be strongest for political news.

To test these hypotheses, we designed an experiment (n=85) that combines participant surveys with the tracking of eye movements when exposed to news items on Facebook. Combining self-reported data with observational measures exposes the flaws of user self-assessment of social media behavior. Participants are exposed to 20 posts from five news genres that were issued to the Facebook page of the Swedish tabloid Aftonbladet in 2017, each post associated with a manipulated comment field. Our manipulation was to create two types of comment pairs: one pair in which commenters agreed with one another, the other pair where the second comment came in disagreement with the first. To increase the ecological validity of the experiment, we use actual news items and comments from Aftonbladet’s Facebook page. Preliminary results confirm that disagreement in comments increases users’ likelihood to read a news item, but decreases their likelihood to share across their own social networks. This finding is particularly strong for the political news items in our sample.

Our research makes a contribution to the literature on social media news sharing and helps understand the consequences of incivility for the dissemination of news on Facebook. We also discuss the implications of using real Facebook data, as opposed to strictly controlled designs, in future experimental designs.

Authors

©2020 All Academic, Inc.   |   Privacy Policy