Five years of accomplishment from the Behavioral Sciences project helped inform the 2014 book, Agent_Zero: Toward Neurocognitive Foundations for Generative Social Science. To view a recording of the author discussing his work, please see the Public Health Dynamics Seminars Archive.
Computational and mathematical models play a central role in epidemiology and public policy, including national pandemic flu policy and other areas of infectious disease preparedness. Many current infectious disease models, however, suffer a crucial shortcoming: human behavioral dimensions are either absent entirely or are severely underdeveloped. The purpose of the Behavior Modeling project is to fundamentally deepen behavorial modeling by incorporating cognitive psychology and neuropsychology into agent behavior. Our project takes the following into consideration when building models:
• Personal health behaviors (e.g.; hand washing, following vaccination recommendations, restricting travel, keeping children out of school, wearing a facemask) play a critical role in the transmission of infectious diseases.
• Health behavior theories (e.g.; the Health Belief Model, Trans-theoretical Model of Behavior Change, or Theory of Planned Behavior) specify underlying motivational and perceptual factors (referred to as theoretical constructs) that determine the likelihood that an individual will adopt a behavior.
• Social factors (e.g.; social influence, disparities, and distrust) constitute the essential context in which decisions are made. The structure of social networks is crucial both in guiding decisions made and in the wider impact of those decisions.
Agent-based modeling lends itself well to behavioral analysis. An agent-based model (ABM) is a computational method developed with the intent of simulating or reproducing the behaviors of autonomous agents within a system. Equilibrium is a central factor in most computational modeling research. Agent-based modeling, however, consists of dynamically-interacting, rule-based agents. Using simple rules, these models can result in extremely complex and emergent behaviors. The project makes a unique contribution by deploying more advanced cognitive agents in model at the USA and global scales. The results can inform high-level decision makers, policy developers, and public health professionals.
- Modeled effects of altruism on vaccination decisions (Shim E, Chapman GB, Townsend JP, Galvani AP. The influence of altruism on influenza vaccination decisions. J R Soc Interface. 2012 Sep 7;9(74):2234-43. doi: 10.1098/rsif.2012.0115. Epub 2012 Apr 11. PubMed PMID: 22496100; PubMed Central PMCID: PMC3405754).
- Pioneered and published methods for using survey research data to parameterize models (Durham DP, Casman EA, Albert SM. Deriving behavior model parameters from survey data: self-protective behavior adoption during the 2009-2010 influenza A(H1N1) pandemic. Risk Anal. 2012 Dec;32(12):2020-31. doi: 10.1111/j.1539-6924.2012.01823.x. Epub 2012 May 7. PubMed PMID: 22563796; PubMed Central PMCID: PMC3755610).
- Developed new methods for automatically detecting trending health-related topics by mining Twitter for emotions, attitudes, and behaviors related to emerging infections [Parker J, Wei Y, Yates A, Frieder O, Goharian N. A Framework for Detecting Public Health Trends with Twitter. ASONAM 2013 (IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining); 2013 August 28; Niagara Falls, Canada].
- Used FRED agent-based model to analyze to impact of paid sick day policies on the spread of epidemics (Kumar S, Grefenstette JJ, Galloway D, Albert SM, Burke DS. Policies to reduce influenza in the workplace: impact assessments using an agent-based model. Am J Public Health. 2013 Aug;103(8):1406-11. doi: 10.2105/AJPH.2013.301269. Epub 2013 Jun 13. PubMed PMID: 23763426; PubMed Central PMCID: PMC3893051).
- Introduced large-scale cognitively credible models in epidemic forecasting, design of containment strategies, and effective risk communication [Agent Zero: Toward Neurocognitive Foundations for Generative Social Science (Epstein, 2013)]