Default image of project.

The Generalizability and Replicability of Twitter Data for Population Research

Social media data have the potential to track phenomena in real time, such as percentage of the population fearful in the minutes after a disaster or terrorist event, or the degree of anger immediately after the announcement of a jury verdict in a highly publicized case. Social media has advantages for researchers that other data collection methods do not have. This project will analyze how the application of survey weighting can rebalance samples of Twitter data, and assesses how well this rebalancing will allow valid generalizations about population behaviors. This project will evaluate the extent to which Twitter users represent or misrepresent the population across different demographic groups and test the feasibility of developing weights that, when applied to Twitter data, make the results more representative of the underlying population. Particular emphasis is given on how properly weighted twitter data can be used in migration relate research. 

Funding Resource: National Science Foundation

Team Members: Guangqing Chi (PI), Jennifer Van Hook, Eric Plutzer, Junjun Yin, and Heng Xu

Default image of project.

Big Data and Network Analysis of Children’s Health

Decades of research suggest that neighborhood socioeconomic disadvantage increases children's health risk. This project seeks to address two major weaknesses in conventional neighborhood effects research and interventions: a) the assumption that residential neighborhoods function independently of each other—ignoring that risk factors in areas where people work, learn, and play away from home may interact with residential factors; and b) as importantly, insufficient understanding of neighborhood effects mechanisms and heterogeneity in effects. I propose a research and training program that will enable me to learn, use, and adapt recent advancements in Big Data analytics. I plan to model hidden interdependencies among individuals and neighborhoods and operationalize mechanisms of neighborhood effects by drawing on multiple large datasets (demographic, geospatial, networks, population flows), with several hundred million observations across multiple states, cities, and years, and match them to locally and nationally representative restricted survey data.

Funding Resource: Eunice Kennedy Shriver National Institute of Child Health and Human Development

Team Members: Corina Graif (PI)

Default image of project.

Are We More Willing to Speak Up through Mobile Phones? A Comparison of Desktop vs. Mobile Political Sharing on Facebook

The small screen size and rushed nature of mobile phone use create bandwidth limitations for users, leading them to be less deliberate and more spontaneous in their online interactions. Does this mean they will be more honest and willing to speak their mind when using social media via mobile phone compared to a desktop computer? We seek to answer this question by investigating the effect of technological devices (mobile phone vs. desktop) on users’ sharing of political content on Facebook. Given that users interact with mobile devices in a spontaneous and intimate way, we hypothesize that users are more likely to speak up through mobile devices on social media, regardless of their political beliefs.

Funding Resource: Social Science Research Council

Team Members: Shyam Sundar (PI), Guangqing Chi, and Junjun Yin

Default image of project.

Pursuing Opportunities for Long-term Arctic Resilience for Infrastructure and Society (POLARIS)

Alaskan coastal Indigenous communities face severe, urgent, and complex social and infrastructural challenges resulting from environmental changes. However, the magnitude and significance of impacts are unclear; as is how local communities will respond to resulting disruptions and disasters. The Pursuing Opportunities for Long-term Arctic Resilience for Infrastructure and Society (POLARIS) project investigates how interconnected environmental stressors and infrastructure disruptions are affecting coastal Arctic Alaskan communities and identifies the important assets (social, environmental, infrastructural, institutional) to help them adapt and become more resilient to climate-related changes. The POLARIS project has identified three convergent research pillars to help communities adapt: environmental hotspots of disruption to communities and infrastructure, food in complex adaptive systems, and migration and community relocation. Research will integrate the pillars where system responses and uncertainties will be predicted under several socio-environmental scenarios.

Funding Resource: National Science Foundation

Team Members: Guangqing Chi (PI), Davin Holen, Ann Tickamyer, Lance Howe, Chris Maio, Kathleen Halvorsen, Erica Smithwick, Kathleen Hill, Anne Jensen, Bronwen Powell, Todd Radenbaugh, Junjun Yin, and Qiujie Zheng

Default image of project.

Pennsylvania Rural Health Model

This project uses the data analysis and GIS services of the CSA team to look at the spatial relationship between the social determinants of health and healthcare facilities within Pennsylvania. The CSA team provides results of three main areas of focus: geospatial mapping of accessibility for the participating hospitals, social and health care determinants of the opioid epidemic in rural Pennsylvania, and utilization of big data to analyze vaccination attitudes in the participating hospitals’ catchment areas.

Funding Resource: U.S. Department of Health and Human Services, Centers for Medicare & Medicaid Services

Team Members: Lisa Davis and Chris Hollenbeak (co-PIs)

Default image of project.

Regional Differences in Opinions Toward Climate Change Among Twitter Users

We aim to study how Twitter, as a social media platform, captures the regional differences of users in climate change conversations. We will tackle three major components. First, we construct a twitter-driven regional opinion heat map regarding attitudes towards climate change. Second, we study how differences in location, time, and occurrences of climate-related disasters shape opinions toward climate change. Third, we want to use out data to create a healthier and more transparent environment for public discussion of climate change.

Funding Resource: PSU Social Science Research Institute, Institute for CyberScience, and College of Information Science and Technology

Team Members: Ting-Hao Huang (PI), Guangqing Chi, and John Yen

Default image of project.

What Can Social Media Sentiment Analysis of Flu Risk and Vaccine Efficacy Tell Us About Regional and Demographic Variations in Flu Vaccination Rates?

This project investigates the predictive potential of social media sentiments, expressed via comments, likes and shares, in determining flu vaccination rates among various regions and demographics in the United States. Despite mounting scientific evidence for the efficacy of flu vaccines, nearly 43 percent of Americans believe that the seasonal flu vaccine can give us the flu. We use large-scale geo-tagged tweets to decipher regional and demographic (especially race) variations in flu-related social media chatter and investigate whether that can be statistically associated with vaccination-related behaviors in those regions and demographic groups. We will identify patterns (combinations of attributes) that correlate to vaccination rates to develop a digital epidemiological framework.

Funding Resource: PSU Social Science Research Institute and Institute for CyberScience

Team Members: Suresh Kuchipudi (PI), Shyam Sundar, and Guangqing Chi

Default image of project.

Explaining Susceptibility to Misinformation in Twitter Using Memory Illusion

The overarching goal of this project is to advance the understanding of how misinformation, in particular, and human behavioral issues, more broadly, can be effectively identified using big data obtained from social media platforms. Specifically, we will integrate advances in data science and key findings from psychological research to investigate how memory illusion can contribute to understanding the dynamics associated with the acquisition of prior information and belief in posterior misinformation on Twitter. Our results will not only have significant intellectual merit in building better machine-based solutions to detect fake news and susceptible users to fake news, but also have far-reaching broader impact on advancing the understanding of human behavior by using big data from naturalistic settings.

Funding Resource: PSU Social Science Research Institute and College of Information Sciences and Technology

Team Members: Dongwon Lee (PI) and Aiping Xiong

Default image of project.

The predictive power of social media engagement on election results: An investigation of bandwagon effects using large-scale geo-tagged tweets

“Bandwagon effect” occurs when individuals tend to favor the majority opinion - “if other people like this, then it must be good for me, too.” We intend to investigate the potential of using social media metrics that quantify the bandwagon associated with political content on Twitter (e.g., number of likes, comments, and shares) to predict spread of social media content and regional election results. We will design a forecasting model that captures the popularity surrounding Twitter content and establish how the bandwagon influences region-specific engagement and outcomes. Specifically, we will examine how geographic proximity and characteristics (e.g., incivility, novelty) of politically-oriented tweets influence bandwagon formation in the weeks before an election and predict regional election outcomes.

Funding Resource: PSU Social Science Research Institute

Team Members: Shyam Sundar (PI) and Nilam Ram