Plastic Surgery and Social Media: Examining Perceptions

2019 
BACKGROUND: Social media play an important role in plastic surgery, yet there are limited studies in the literature to guide plastic surgeons' social networking practices. To address this deficiency and provide further insight, the authors set out to investigate the public's attitude toward plastic surgery using Twitter, a popular social media platform. The authors examined a large body of messages (tweets) related to plastic surgery using novel techniques of natural language processing and sentiment analysis. METHODS: The authors collected over 1 million tweets with the keywords "plastic," "cosmetic," "aesthetic," and "reconstruction" surgery spanning from 2012 to 2016 from the Twitter Gardenhose feed. Using hedonometrics, the authors extracted the average happiness/positivity (havg) of tweets and created word-shift graphs to determine the most influential words. RESULTS: The positivity scores for keywords "plastic," "cosmetic," "aesthetic," and "reconstruction" surgery were 5.72, 6.00, 6.16, and 6.09, respectively. In relation to "plastic," keywords "cosmetic" and "aesthetic" were more positive because they lacked antagonistic words, such as "fake," "ugly," "bad," "fails," or "wrong." The keyword "reconstruction," however, was more positively associated than the term "plastic" because of an increase in positive words, such as "honor," "amazing," "successful," and "respect. CONCLUSIONS: Tweets containing the term "plastic" surgery trended toward negativity, and may be explained by the increase in unfavorable, associative words. Conversely, related terms such as "aesthetic," "cosmetic," and "reconstruction" were more favorably regarded because of the lack of antagonistic words and the presence of supportive words. The authors' results are informative and may serve to guide plastic surgeons' social media practices.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    25
    References
    16
    Citations
    NaN
    KQI
    []