An Analysis of Emotions in Educationally Relevant Tweets

• In Event: Antecedents and Consequences of Emotions in Learning

Abstract

Social networking sites such as Twitter make it possible for anyone in the world to publish a public message on any topic, including educationally relevant topics. The present study demonstrates how this data can be collected, mined, and analyzed. Furthermore, this study provides an investigation of the prevalence and the temporal stability of emotions reported in educationally relevant messages (tweets) posted on Twitter. Almost 5 million English public messages, from around the world, were analyzed for Pekrun’s achievement emotions, and Ekmans’ basic emotions. Results suggest that the most frequent emotion reported was enjoyment, and that emotions fluctuate in specific patterns by the hours of the day, the days of the week, and the months of the year.

Authors

• John Ranellucci, Michigan State University

• François Bouchet, McGill University

• Eric G. Poitras, University of Utah

• Susanne P. Lajoie, McGill University

• Roger Azevedo, North Carolina State University