Affiliation: School of Information and Library Science
Through the lens of the exploratory framework of Digital Trace of Scholarly Acts (DTSA), this dissertation study explored researchers’ activities around scholarly articles on Twitter. Using a mixed-methods design, this study analyzed data collected from a large-scale survey and twenty interviews with researchers on Twitter. The Critical Incident Technique was used as part of the interview study to learn about the full stories behind researchers’ sharing of scholarly articles on Twitter. There were variations in the researcher’s sentiment of opinions on articles they tweeted, retweeted, replied, and liked, based on their demographics. Despite a general positive tendency, researchers’ Twitter activities were associated with different sentiment due to their different perceptions of these activities. Variations were also found in how sharing scholarly articles on Twitter fit into researchers’ scholarly acts workflow with no monolithic pattern. This study contributed to a better understanding of the digital traces left by researchers on Twitter by providing richer descriptions and narratives of their activities. Researchers shared scholarly articles on Twitter for a variety of motivations: networking, promoting, disseminating, commenting, communicating with intended users, acknowledgment, and saving for later reference. The findings particularly shed light on the role of Twitter in communicating research and network building. Investigating the impact of the articles on the researchers led to a better understanding of what types of articles had a higher premium of sharing by researchers on Twitter. Evidence was found to support both the normative theory and the constructivist theory – the categories of impact included connecting, informing, practice-changing, beyond research, and potential impact. However, more than half of the shared articles examined had no impact on the researchers’ own work, indicating that Twitter metrics, even solely based on researchers’ Twitter activities, should not be used as an evaluative metric of the articles shared.