Journal of Information Technology & Politics ( IF 2.710 ) Pub Date : 2021-09-20 , DOI: 10.1080/19331681.2021.1977757 Xiao Wang , Yang Yu
ABSTRACT
Given the sizable viewership of the U.S. presidential debates and the importance of moral sentiments in human behavior, this present investigation examined the relationship between moral framing in the U.S. presidential debates and the changes of moral words and self-transcendent emotions in tweets. For each presidential debate in 2016, we collected tweets on the day of the debate and the day after the debate. A total of 991,835 tweets were coded using the “bag-of-words” approach and supervised machine learning. The results showed that Clinton’s moral framing during the 2016 presidential debates was generally associated with harm/care and fairness in the tweets, whereas Trump’s moral framing was associated with ingroup loyalty and authority in the tweets. Interrupted time series analyses indicated that the effects of the debates on the moral words and self-transcendent emotions in the tweets were limited. Political candidates should aim to induce and then sustain moral sentiments in voters.
中文翻译:
2016年美国总统辩论中道德框架对推文中道德判断和自我超越情绪的影响
摘要
鉴于美国总统辩论的大量收视率以及道德情绪在人类行为中的重要性,本次调查研究了美国总统辩论中的道德框架与推文中道德话语和自我超越情绪的变化之间的关系。对于 2016 年的每一场总统辩论,我们都会在辩论当天和辩论后的第二天收集推文。使用“词袋”方法和监督机器学习对总共 991,835 条推文进行了编码。结果表明,克林顿在 2016 年总统辩论期间的道德框架通常与推文中的伤害/关怀和公平相关,而特朗普的道德框架与推文中的内群体忠诚度和权威相关。中断的时间序列分析表明,辩论对推文中的道德话语和自我超越情绪的影响是有限的。政治候选人应该旨在诱导并维持选民的道德情绪。