Gi Woong Yun1, David Morin2, Sanghee Park3, Claire Youngnyo Joa4, Brett Labbe5, Jongsoo Lim6, Sooyoung Lee7, Daewon Hyun8. 1. Bowling Green State University, School of Media and Communication, Bowling Green, OH 43403, United States. Electronic address: gyun@bgsu.edu. 2. Utah Valley University, Department of Communication, Orem, UT 84058, United States. Electronic address: david.morin@uvu.edu. 3. Bowling Green State University, School of Media and Communication, Bowling Green, OH 43403, United States. Electronic address: spark@bgsu.edu. 4. Bowling Green State University, School of Media and Communication, Bowling Green, OH 43403, United States. Electronic address: yjoa@bgsu.edu. 5. Bowling Green State University, School of Media and Communication, Bowling Green, OH 43403, United States. Electronic address: blabbe@bgsu.edu. 6. Sejong University, Department of Communication Arts, Seoul, Republic of Korea. Electronic address: jslim123@sejong.ac.kr. 7. Sogang University, School of Communication, Seoul, Republic of Korea. Electronic address: sooyoung@sogang.ac.kr. 8. Sogang University, School of Communication, Seoul, Republic of Korea. Electronic address: dhyun@sogang.ac.kr.
Abstract
OBJECTIVES: This paper has two objectives. First, it categorizes the Twitter handles tweeted flu related information based on the amount of replies and mentions within the Twitter network. The collected Twitter accounts are categorized as media, health related individuals, organizations, government, individuals with no background with media or medical field, in order to test the relationship between centrality measures of the accounts and their categories. The second objective is to examine the relationship between the importance of the Twitter accounts in the network, centrality measures, and specific characteristics of each account, including the number of tweets and followers as well as the number of accounts followed and liked. METHODS: Using Twitter search network API, tweets with "flu" keyword were collected and tabulated. Network centralities were calculated with network analysis tool, NodeXL. The collected Twitters accounts were content analyzed and categorized by multiple coders. RESULTS: When the media or organizational Twitter accounts were present in the list of important Twitter accounts, they were highly effective disseminating flu-related information. Also, they were more likely to stay active one year after the data collection period compared to other influential individual accounts. CONCLUSIONS: Health campaigns are recommended to focus on recruiting influential Twitter accounts and encouraging them to retweet or mention in order to produce better results in disseminating information. Although some individual social media users were valuable assets in terms of spreading information about flu, media and organization handles were more reliable information distributors. Thus, health information practitioners are advised to design health campaigns better utilizing media and organizations rather than individuals to achieve consistent and efficient campaign outcomes. Published by Elsevier Ireland Ltd.
OBJECTIVES: This paper has two objectives. First, it categorizes the Twitter handles tweeted flu related information based on the amount of replies and mentions within the Twitter network. The collected Twitter accounts are categorized as media, health related individuals, organizations, government, individuals with no background with media or medical field, in order to test the relationship between centrality measures of the accounts and their categories. The second objective is to examine the relationship between the importance of the Twitter accounts in the network, centrality measures, and specific characteristics of each account, including the number of tweets and followers as well as the number of accounts followed and liked. METHODS: Using Twitter search network API, tweets with "flu" keyword were collected and tabulated. Network centralities were calculated with network analysis tool, NodeXL. The collected Twitters accounts were content analyzed and categorized by multiple coders. RESULTS: When the media or organizational Twitter accounts were present in the list of important Twitter accounts, they were highly effective disseminating flu-related information. Also, they were more likely to stay active one year after the data collection period compared to other influential individual accounts. CONCLUSIONS: Health campaigns are recommended to focus on recruiting influential Twitter accounts and encouraging them to retweet or mention in order to produce better results in disseminating information. Although some individual social media users were valuable assets in terms of spreading information about flu, media and organization handles were more reliable information distributors. Thus, health information practitioners are advised to design health campaigns better utilizing media and organizations rather than individuals to achieve consistent and efficient campaign outcomes. Published by Elsevier Ireland Ltd.
Keywords:
Centrality; Flu; Health information campaign; Media; Social media; Social network analysis; Twitter
Authors: Erika Bonnevie; Sarah D Rosenberg; Caitlin Kummeth; Jaclyn Goldbarg; Ellen Wartella; Joe Smyser Journal: PLoS One Date: 2020-10-16 Impact factor: 3.240