IUDs in the Twittersphere: Sentiment Analysis of IUD-Related Tweets [23F]. Issue 1 (May 2017)
- Record Type:
- Journal Article
- Title:
- IUDs in the Twittersphere: Sentiment Analysis of IUD-Related Tweets [23F]. Issue 1 (May 2017)
- Main Title:
- IUDs in the Twittersphere
- Authors:
- Jung, Christina
Richter, Felix
Lunde, Britt
Chen, Katherine - Abstract:
- Abstract : INTRODUCTION: Twitter has allowed for rapid sharing of user-created content. Researchers have begun to study sentiment in Twitter by building a word lexicon scored for valence from negative to positive sentiment, creating a method of validated sentiment analysis. However, few studies have looked at Twitter and health—much less, in relation to contraception and women's health. METHODS: Publicly posted tweets from January 1, 2007 to January 1, 2016 with hashtags referencing IUD and brands Mirena, Skyla, and Liletta were separated into the following sources: 1) "individual", 2) "medical provider", 3) "health organization", and 4) "legal organization" based on username, Twitter handle, and full name. Tweet content was interpreted with a validated sentiment analysis, and compared across source and brand. RESULTS: We found that Tweets by individuals were significantly more positive than Tweets by healthcare organizations (p < 0.005). Tweets by law organizations were significantly more negative than all other sources (p < 0.0005) Tweets about IUD brands across all sources spiked in number at the time of each device's release into the market. Tweets about the newer IUD brands, Skyla and Liletta, were more positive than those about the Mirena (p < 0.0005). CONCLUSION: Our study shows how Twitter data can reflect temporal and sentiment patterns in lay perception of and reaction to IUDs over time. Studying how information about contraception is shared on social mediaAbstract : INTRODUCTION: Twitter has allowed for rapid sharing of user-created content. Researchers have begun to study sentiment in Twitter by building a word lexicon scored for valence from negative to positive sentiment, creating a method of validated sentiment analysis. However, few studies have looked at Twitter and health—much less, in relation to contraception and women's health. METHODS: Publicly posted tweets from January 1, 2007 to January 1, 2016 with hashtags referencing IUD and brands Mirena, Skyla, and Liletta were separated into the following sources: 1) "individual", 2) "medical provider", 3) "health organization", and 4) "legal organization" based on username, Twitter handle, and full name. Tweet content was interpreted with a validated sentiment analysis, and compared across source and brand. RESULTS: We found that Tweets by individuals were significantly more positive than Tweets by healthcare organizations (p < 0.005). Tweets by law organizations were significantly more negative than all other sources (p < 0.0005) Tweets about IUD brands across all sources spiked in number at the time of each device's release into the market. Tweets about the newer IUD brands, Skyla and Liletta, were more positive than those about the Mirena (p < 0.0005). CONCLUSION: Our study shows how Twitter data can reflect temporal and sentiment patterns in lay perception of and reaction to IUDs over time. Studying how information about contraception is shared on social media platforms like Twitter can help us better understand the context in which patients make decisions about their health. … (more)
- Is Part Of:
- Obstetrics and gynecology. Volume 129:Issue 1(2017)
- Journal:
- Obstetrics and gynecology
- Issue:
- Volume 129:Issue 1(2017)
- Issue Display:
- Volume 129, Issue 1 (2017)
- Year:
- 2017
- Volume:
- 129
- Issue:
- 1
- Issue Sort Value:
- 2017-0129-0001-0000
- Page Start:
- Page End:
- Publication Date:
- 2017-05
- Subjects:
- Obstetrics -- Periodicals
Gynecology -- Periodicals
618 - Journal URLs:
- http://journals.lww.com/greenjournal/pages/default.aspx ↗
http://journals.lww.com ↗ - DOI:
- 10.1097/01.AOG.0000514852.66590.1f ↗
- Languages:
- English
- ISSNs:
- 0029-7844
- Deposit Type:
- Legaldeposit
- View Content:
- Available online (eLD content is only available in our Reading Rooms) ↗
- Physical Locations:
- British Library DSC - 6208.200000
British Library DSC - BLDSS-3PM
British Library HMNTS - ELD Digital store - Ingest File:
- 4529.xml