人气
社会化媒体
互联网隐私
杠杆(统计)
假新闻
情感(语言学)
广告
计算机科学
心理学
万维网
业务
社会心理学
沟通
机器学习
作者
Shuting Wang,Min‐Seok Pang,Paul A. Pavlou
标识
DOI:10.25300/misq/2022/16296
摘要
Social media platforms, such as Facebook, Instagram, and Twitter, are combating the spread of fakennews by developing systems that allow their users to report fake news. However, it remains unclear whether these reporting systems that harness the “wisdom of the crowd” are effective. Notably, concerns have been raised that the popularity of videos may hamper users’ reporting of fake news. The persuasive power of videos may render fake news more deceptive and less likely to be reported in practice. However, this is neither theoretically nor empirically straightforward, as videos not only affect users’ ability to detect fake news, but also impact their willingness to report and their engagement (i.e., likes, shares, and comments) which would further spread fake news. Using a unique dataset from a leading social media platform, we empirically examine how including a video in a fake news post affects the number of users reporting the post to the platform. Our results indicate that including a video significantly increases the number of users reporting the fake news post to the social media platform. Additionally, we find that the sentiment intensity of the fake news text content, especially when the sentiment is positive, attenuates the effect of including a video. Randomized experiments and a set of mediation analyses are included to uncover the underlying mechanisms. We contribute to the information systems literature by examining how social media platforms can leverage their users to report fake news, and how different formats (e.g., videos and text) of fake news interact to influence users’ reporting behavior. Social media platforms that seek to leverage the “wisdom of the crowd” to combat the proliferation of fake news should consider both the popularity of videos and the role of text sentiment in fake news to adjust their strategies.
科研通智能强力驱动
Strongly Powered by AbleSci AI