Sentiments towards Emergency Remote Teaching on Twitter: A Longitudinal Comparative Sentiment Analysis

Authors : Secil CASKURLU, Serkan AYVAZ, Kadir KOZAN

Pages : 1-21

Doi:10.54535/rep.1206924

View : 15 | Download : 14

Publication Date : 2023-04-30

Article Type : Research Paper

Abstract :This longitudinal and comparative study investigated people’s sentiments toward emergency remote teaching in tweets posted in two different languages from January 10 to August 16 2021 when mass vaccinations started and continued. The results indicated that English tweets insert ignore into journalissuearticles values(a); included more positive sentiments towards emergency remote teaching; insert ignore into journalissuearticles values(b); were more supportive and motivating; and insert ignore into journalissuearticles values(c); focused on topics related to education, online education, and English as a second or foreign language. However, Turkish tweets insert ignore into journalissuearticles values(a); included more similar amounts of neutral and positive sentiments; insert ignore into journalissuearticles values(b); involved politics and government-related content; and insert ignore into journalissuearticles values(c); touched on topics related to preschool education, ministry of national education and the e-school system used during the pandemic. Lastly, compared to positive and neutral sentiments, there were fewer negative sentiments in tweets in both languages suggesting that people got used to emergency remote teaching over time. In other words, despite any ongoing issues, people’s reactions to emergency remote teaching on Twitter improved and became either more neutral or positive in a year or so, which implies that increasing optimism due to vaccinations during sudden health crises may calibrate people’s sentiments towards compulsory solutions such as emergency remote teaching.
Keywords : Comparative sentiment analysis, emergency online education, Twitter, topic modeling, text mining, emergency online education, Twitter, topic modeling, text mining