Investigating Twitter mentions of research papers as altmetric indicators

• Main Author: Tint Hla Hla Htoo
• Other Authors: Na Jin Cheon
• Format: Thesis-Doctor of Philosophy
• Language: English
• Published: Nanyang Technological University 2023
• Subjects: Social sciences::Communication; Library and information science
• Online Access: https://hdl.handle.net/10356/168524
• Institution: Nanyang Technological University

During the past decade, Twitter has emerged as the social media platform with the most extensive coverage of research papers and a potentially useful source of altmetrics in revealing the societal impacts of research. However, it is still unclear what general conclusions can be drawn about research articles that are frequently tweeted and what Twitter mentions (number of tweets or tweeters mentioning research papers) reflect. This research investigated Twitter mentions in three areas and revealed new knowledge founded on theories to contribute to our understanding of Twitter mentions as altmetric indicators. The first study, Study One, in this dissertation investigated the validity of Twitter mentions as crowdsourced indicators of research significance. Considering retweets as a manifestation of social influence, the study first explored whether tweeting research papers signifies the presence of either the collective intelligence phenomenon, known as the wisdom of the crowd effect, or a herd-like behavior for a dataset of COVID-19 papers from the Lancet. The study then used the MAIN (Modality, Agency, Interactivity, and Navigability) model to examine the nature of the influence involved in retweeting. Next, to find what Twitter mentions of research papers reflect, Study Two investigated characteristics of research papers that may cause some papers to be tweeted more. Using news value theory, this study was the first in altmetrics research to investigate news value (or newsworthiness), measured by news factors, as a key characteristic of research papers associated with tweet counts or tweeter counts. It was found that all the news factors examined (importance, controversy, elite nations, elite persons, scale, and news prominence) were associated with higher tweet/tweeter counts. Since these factors can theoretically be considered to be general human selection criteria or relevance indicators, the findings give new evidence that Twitter mentions may be indicators of general relevance to members of society. Study Two also provided a new understanding of the strong positive relationship between media coverage and Twitter mentions received by the papers. Instead of news coverage bringing more tweets or the other way around (journalists noticing highly tweeted articles and writing about them), results suggested that research papers with news value could attract Twitter mentions regardless of the news mentions they receive. Therefore, in the presence of news factors, it was not the news mentions that contributed to the Twitter mentions, but it was more likely the news value of the papers that attracted both the Twitter mentions and the news mentions. Finally, this research provided new evidence in altmetrics research that the demographic representation of tweeters tweeting research papers was significantly different from that of the general population. Racial representation in the retweeter population was found to closely follow the pattern in the retweeted original tweeter population, indicating the possible role of original tweeters in the popularity of research papers and highlighting the need for more attention in this area in research on Twitter mentions. As for differences in motivations among different user types, the main difference was found in the Promotion category. As academic organization users were found to promote or self-cite their papers in their tweets, the results of this research indicated that it is worth exploring how self-promotion contributes to overall tweet counts received by the articles. This research also confirmed that Twitter mentions reflected both academic and non-academic impacts. In conclusion, this research contributed to the altmetrics literature by demonstrating the applicability of both new and established concepts from social theories in identifying issues surrounding Twitter mentions and in getting a deeper understanding of them. As altmetrics research is a new context in which these concepts were applied for the first time, this research also made important contributions to these theories by showing their applicability in explaining this new phenomenon.