0261 Discourse Around Sleep, Performance, and Work in the U.S. Print Media: 1983-2017

Abstract Introduction This computational social science study examined trends in sleep discourse in major U.S. newspapers from 1983 to 2017. If societies undergo change over time, so too might sleep norms and expectations. The purpose of this report is to document and investigate the possible shifting significance of sleep over this time period. Methods Leveraging the large-scale digital archiving of newspapers from the 1980s, news articles from major U.S. newspapers were acquired from ProQuest (1983-2017). Articles were searched by the following criteria: “sleep” appeared in the title or abstract, from the New York Times, Washington Post, Chicago Tribune, USA Today, the Wall Street Journal, the Orange County Register, and Los Angeles Times. A 70-topic structural topic model was run on the corpus containing 25,766 documents. Results The data show tension between work and sleep but also show attempts at ‘customization’ which emphasizes the importance of protecting sleep for performance and productivity. Further, the link between poor sleep and adverse health outcomes is increasingly emphasized. Naps emerged as a significant concept for discussion in relation to work and night-time sleep. Conclusion An analysis of U.S. print media from 1983 to 2017 suggests attempts at ‘customization’, that is, towards the view of sleep as beneficial for performance and productivity. One rationale for customization may be that since cultural attitudes towards work are unlikely change, and work is of prime importance, sleep must be framed as complementary to work. America’s relationship with sleep and work could be viewed as an evolving discussion on how to square the necessity of sleep for optimal performance with the fact that an hour more for sleep is an hour less for work or leisure. Support (if any) Training and research: Emory University, Department of Sociology; Martin Levinson departmental grant to attend the Inter-university Consortium for Political and Social Research (ICPSR) summer course “Machine Learning for the Analysis of Text as Data” (2017). Writing: 5R01HL152453.