Everyone Has an Opinion and there are No Rules: How U.S. Journalists Articulate the Impact of Social Media on Journalism

This is the authors' abstract. We don't add key points for this paper.

ABSTRACTThis study examines how newsworkers make sense of the impact of social media on journalism. We draw on the concept of metajournalistic discourse as a lens through which to analyze the web of positive and negative meanings evoked by journalists when assessing the industry impact of social media. We analyze data from a national survey of U.S. journalists (N = 1272), combining a computational analysis of responses to open-ended questions with qualitative analysis. Findings illustrate the rich complexity of negative discourse about social media in journalism, comprising an array of articulations linking social media to the delegitimization of news.KEYWORDS: Metajournalistic discourseplatformssocial mediasurveystructural topic modelingjournalists Disclosure StatementNo potential conflict of interest was reported by the author(s).Notes1 We removed 271 participants who did not provide a response or wrote less than five words for the negative or positive impact questions.2 As a check of response validity, we conducted a sentiment analysis examining whether the two open-ended questions produced responses that differed in sentiment. Results confirm that responses to the negative impact question contained more negative language than responses to the positive impact question; likewise, the positive impact question elicited more positive language than response to the negative impact question. For more details about this analysis, see the supplemental Appendix.3 Specifically, we test whether differences at the individual-level (e.g., age, education) and organizational-level (e.g., organization type, size) predict the use of certain negative and positive themes. Due to space limitations, these findings are presented in the supplemental Appendix, Tables A2 and A3.4 Due to unequal topic sizes, two slight modifications were made to the sampling strategy. The following probability cuts were used: 1) delivering the news farther and faster 0.55 or above; 2) a more equitable future of news 0.50 or above; and 3) new pathways for news 0.45 or above.5 This example also reflects a tendency we found in our data for longer positive responses to become increasingly more negative in sentiment. See the supplemental Appendix and Figure A1 for more information.