Affective responses to uncertain real-world outcomes: Sentiment change on Twitter
Affective responses to uncertain real-world outcomes: Sentiment change on Twitter
复制标题
DOI：
10.1371/journal.pone.0212489
复制
发表时间：
2019-02-27
期刊：
影响因子：
3.7
通讯作者：
Walasek, Lukasz
中科院分区：
文献类型：
--
作者：
Bhatia, Sudeep;Mellers, Barbara;Walasek, Lukasz
我们使用来自Twitter.com的数据来研究对不确定结果的影响和期望之间的相互作用。在两项研究中，我们获得了关于2014年美国参议院选举候选人的推文，以及关于2014/2015年NFL赛季NFL球队的推文。我们选择这些事件是因为a）它们的结果高度不确定，B）它们吸引了大量的关注，并在社交媒体上的传播中占有重要地位。我们还从政治预测和体育博彩网站获得了对事件的先验期望。使用这种准实验设计，我们发现，意想不到的事件与更强烈的影响比预期的事件。此外，对于低于预期的结果，预期的影响要大于超出预期的结果。我们的研究结果与心理科学的基本原理是一致的，例如经验情感中的参照依赖性。我们讨论如何自然发生的在线数据可以用来测试心理预测和开发新的心理见解。
We use data from Twitter.com to study the interplay between affect and expectations about uncertain outcomes. In two studies, we obtained tweets about candidates in the 2014 US Senate elections and tweets about National Football League (NFL) teams in the 2014/2015 NFL season. We chose these events because a) their outcomes are highly uncertain and b) they attract a lot of attention and feature heavily in the communication on social media. We also obtained a priori expectations for the events from political forecasting and sport betting websites. Using this quasi-experimental design, we found that unexpected events are associated with more intense affect than expected events. Moreover, the effect of expectations is larger for outcomes that fall below expectations than outcomes that exceed expectations. Our results are consistent with fundamental principles in psychological science, such as reference-dependence in experienced affect. We discuss how naturally occurring online data can be used to test psychological predictions and develop novel psychological insights.