Assessing the Russian Internet Research Agency's impact on the political attitudes and behaviors of American Twitter users in late 2017.

There is widespread concern that Russia and other countries have launched social-media campaigns designed to increase political divisions in the United States. Though a growing number of studies analyze the strategy of such campaigns, it is not yet known how these efforts shaped the political attitudes and behaviors of Americans. We study this question using longitudinal data that describe the attitudes and online behaviors of 1,239 Republican and Democratic Twitter users from late 2017 merged with nonpublic data about the Russian Internet Research Agency (IRA) from Twitter. Using Bayesian regression tree models, we find no evidence that interaction with IRA accounts substantially impacted 6 distinctive measures of political attitudes and behaviors over a 1-mo period. We also find that interaction with IRA accounts were most common among respondents with strong ideological homophily within their Twitter network, high interest in politics, and high frequency of Twitter usage. Together, these findings suggest that Russian trolls might have failed to sow discord because they mostly interacted with those who were already highly polarized. We conclude by discussing several important limitations of our study-especially our inability to determine whether IRA accounts influenced the 2016 presidential election-as well as its implications for future research on social media influence campaigns, political polarization, and computational social science.

Assessment of ownership of smart devices and the acceptability of digital health data sharing.

Smart portable devices- smartphones and smartwatches- are rapidly being adopted by the general population, which has brought forward an opportunity to use the large volumes of physiological, behavioral, and activity data continuously being collected by these devices in naturalistic settings to perform research, monitor health, and track disease. While these data can serve to revolutionize health monitoring in research and clinical care, minimal research has been conducted to understand what motivates people to use these devices and their interest and comfort in sharing the data. In this study, we aimed to characterize the ownership and usage of smart devices among patients from an expansive academic health system in the southeastern US and understand their willingness to share data collected by the smart devices. We conducted an electronic survey of participants from an online patient advisory group around smart device ownership, usage, and data sharing. Out of the 3021 members of the online patient advisory group, 1368 (45%) responded to the survey, with 871 female (64%), 826 and 390 White (60%) and Black (29%) participants, respectively, and a slight majority (52%) age 58 and older. Most of the respondents (98%) owned a smartphone and the majority (59%) owned a wearable. In this population, people who identify as female, Hispanic, and Generation Z (age 18-25), and those completing higher education and having full-time employment, were most likely to own a wearable device compared to their demographic counterparts. 50% of smart device owners were willing to share and 32% would consider sharing their smart device data for research purposes. The type of activity data they are willing to share varies by gender, age, education, and employment. Findings from this study can be used to design both equitable and cost-effective digital health studies, leveraging personally-owned smartphones and wearables in representative populations, ultimately enabling the development of equitable digital health technologies.

Reducing political polarization in the United States with a mobile chat platform.

Do anonymous online conversations between people with different political views exacerbate or mitigate partisan polarization? We created a mobile chat platform to study the impact of such discussions. Our study recruited Republicans and Democrats in the United States to complete a survey about their political views. We later randomized them into treatment conditions where they were offered financial incentives to use our platform to discuss a contentious policy issue with an opposing partisan. We found that people who engage in anonymous cross-party conversations about political topics exhibit substantial decreases in polarization compared with a placebo group that wrote an essay using the same conversation prompts. Moreover, these depolarizing effects were correlated with the civility of dialogue between study participants. Our findings demonstrate the potential for well-designed social media platforms to mitigate political polarization and underscore the need for a flexible platform for scientific research on social media.

Protecting the integrity of survey research.

Although polling is not irredeemably broken, changes in technology and society create challenges that, if not addressed well, can threaten the quality of election polls and other important surveys on topics such as the economy. This essay describes some of these challenges and recommends remediations to protect the integrity of all kinds of survey research, including election polls. These 12 recommendations specify ways that survey researchers, and those who use polls and other public-oriented surveys, can increase the accuracy and trustworthiness of their data and analyses. Many of these recommendations align practice with the scientific norms of transparency, clarity, and self-correction. The transparency recommendations focus on improving disclosure of factors that affect the nature and quality of survey data. The clarity recommendations call for more precise use of terms such as "representative sample" and clear description of survey attributes that can affect accuracy. The recommendation about correcting the record urges the creation of a publicly available, professionally curated archive of identified technical problems and their remedies. The paper also calls for development of better benchmarks and for additional research on the effects of panel conditioning. Finally, the authors suggest ways to help people who want to use or learn from survey research understand the strengths and limitations of surveys and distinguish legitimate and problematic uses of these methods.

The dynamics of health care opinion, 2008-2010: partisanship, self-interest, and racial resentment.

Recent debate over the Patient Protection and Affordable Care Act highlights the extent of party polarization in Washington. While the partisan divide on this issue is stark among political elites, it is less clear how the mass electorate has responded to this divisive conflict. In this article we examine individual-level dynamics in health care attitudes between 2008 and 2010. We find partisan attachments and self-interests strongly predict change in health care attitudes, with Republicans growing more opposed to universal health insurance between 2008 and 2010, and those personally worried about medical expenses less likely to abandon support. We find, however, that the effect of partisanship is moderated by self-interest, with strong Republicans significantly less likely to switch to opposition if they were personally worried about medical expenses. Finally, we find that health care policy preferences, already tinged with racial attitudes in 2008, became increasingly so by 2010.

Leveraging Auxiliary Information on Marginal Distributions in Nonignorable Models for Item and Unit Nonresponse.

Often, government agencies and survey organizations know the population counts or percentages for some of the variables in a survey. These may be available from auxiliary sources, for example, administrative databases or other high quality surveys. We present and illustrate a model-based framework for leveraging such auxiliary marginal information when handling unit and item nonresponse. We show how one can use the margins to specify different missingness mechanisms for each type of nonresponse. We use the framework to impute missing values in voter turnout in a subset of data from the U.S. Current Population Survey (CPS). In doing so, we examine the sensitivity of results to different assumptions about the unit and item nonresponse.

The National Sleep Foundation's Sleep Satisfaction Tool.

OBJECTIVES: The National Sleep Foundation (NSF) sought to test, refine, and add statistical rigor to its previously described provisional Sleep Satisfaction Tool (SST). The tool assesses the general population's sleep satisfaction. DESIGN: In 2017, NSF created a provisional tool through systematic literature review and an expert consensus panel process. This tool was expanded, refined, and tested through an open-ended survey, 2 rounds of cognitive testing, and a national survey of a random sample of Internet users (aged 18-90). Factor analysis and final consensus panel voting produced the robust SST. RESULTS: The exploratory, open-ended surveying for identifying additional factors important to the public led to question formulation around mind relaxation. Cognitive testing yielded significant refinement to question and response option formatting. Factor analysis of questions from field testing indicated loading on one construct identified as "sleep satisfaction." The final 9-item SST demonstrated strong reliability and internal validity with overall SST scores of 56/100 (higher scores indicating greater sleep satisfaction). Individual SST item mean scores ranged from 39 to 66, and overall SST scores varied substantially across demographic groups. CONCLUSIONS: NSF used a series of development and validation tests on its provisional SST, producing a novel and reliable research tool that measures the general population's sleep satisfaction. The SST is a short, reliable, nonclinical assessment that expands the set of tools available to researchers that implements the individual, social, and environmental factors related to sleep satisfaction. Further research will explore refined scoring methods along with factor weighting and use within different populations.

The National Sleep Foundation's Sleep Health Index.

OBJECTIVES: A validated survey instrument to assess general sleep health would be a useful research tool, particularly when objective measures of sleep are not feasible. Thus, the National Sleep Foundation spearheaded the development of the Sleep Health Index (SHI). DESIGN: The development of the SHI began with a task force of experts who identified key sleep domains and questions. An initial draft of the survey was created and questions were refined using cognitive testing and pretesting. The resulting 28-question survey was administered via random-sample telephone interviews to nationally representative samples of adults in 2014 (n=1253) and 2015 (n=1250). These data were combined to create the index. A factor analysis linked 14 questions to 3 discrete domains: sleep quality, sleep duration, and disordered sleep. These were assembled as sub-indices, then combined to form the overall SHI, with scores ranging from 0 to 100 (higher score reflects better sleep health). RESULTS: Americans earned an overall SHI score of 76/100, with sub-index scores of 81/100 in disordered sleep, 79/100 in sleep duration, and 68/100 in sleep quality. In regression analyses, the strongest independent predictors of sleep health were self-reported stress (ß=-0.26) and overall health (ß=0.26), which were also the strongest predictors of sleep quality (ß=-0.32 and ß=0.27 respectively). CONCLUSIONS: The current 12-item SHI is a valid, reliable research tool that robustly measures 3 separate but related elements of sleep health-duration, quality, and disorders-and assesses the sleep health status of adults in the United States.