Computational Social Science Society of the Americas

CSSSA Webinar Series

We are very excited to host Matthew Koehler and Andrew Collins presenting Validation of Agent-based Models. Validation is the process of determining if a model adequately represents the system under study for the model’s intended purpose. Validation is a critical component in building the credibility of a simulation model with its end-users. Effectively conducting validation can be a daunting task for both novice and experienced simulation developers. Further compounding the difficult task of conducting validation is that there is no universally accepted approach for assessing a simulation. These challenges are particularly relevant to the paradigm of Agent-Based Modeling and Simulation (ABMS) because of the complexity found in these models’ mechanisms and in the real-world situations they attempt to represent. In this presentation, the validation of agent-based models will be introduced, some of the methods and approaches to conducting a validation process will be discussed, and practical tips will be provided. Further to this introduction, a more philosophical discussion about validation will be included. Specifically, a discussion of a new emphasis on thinking about validation in terms of “invalidation.” This philosophical view is hoped to be a starting point for a wider discussion on validation at this year’s CSS annual conference in October. This presentation will be at the introductory level of conducting ABMS validation process but should have something for all experience levels. Bios: Dr. Matthew Koehler is computational social scientist at the MITRE Corporation, where he leads the generative social science and computational finance group. During his time at MITRE, Matt has concentrated on decision support using spatially aware agent-based models and simulations, and the analysis and visualization of datasets produced by complex, often social, systems. He has authored several journal articles, peer reviewed conference papers, and book chapters on applied agent-based modeling from tax evasion to combat, and fraud detection to legal institution design. Matt received his AB in Anthropology from Kenyon College, his MPA from Indiana University’s School of Public and Environmental Affairs, his JD from George Washington University’s Law School, and his PhD in Computational Social Science from George Mason University’s Krasnow Institute for Advanced Study, Dept. of Computational Social Science. Matt has taught applied agent-based modeling and data mining at George Mason University, the United States Military Academy, the Università del Salento, and at MITRE’s internal education facility and various government agencies. Matt has also been an active member of the board of the Computational Social Science Society of the Americas for more than a dozen years. Andrew J. Collins, Ph.D., is an associate professor at Old Dominion University in the Department of Engineering Management and Systems Engineering. He has a Ph.D. in Operations Research from the University of Southampton, and his undergraduate degree in Mathematics was from the University of Oxford. He has published over 100 peer-reviewed articles. He has been the Principal Investigator on projects funded to the amount of approximately $5 million. Dr. Collins' interests are the advancement of agent-based modeling (ABM) and the integration of game theory concepts in ABM. has developed several research simulations, including an award-winning investigation into the foreclosure contagion that incorporated social networks. His website and full resume are at www.drandrewjcollins.com.
[CSSSA Webinar] Validation of Agent-based Models – Drs. Matthew Koehler & Andrew Collins
Dr. Arianna Salazar-Miranda presents "the 15-minute city quantified using human mobility data." Amid rising congestion and transport emissions, policymakers are embracing the ‘15-minute city’ model, which envisions neighborhoods where basic needs can be met within a short walk from home. Prior research has primarily examined amenity access without exploring its relationship to behavior. We introduce a measure of local trip behavior using GPS data from 40 million US mobile devices, defining ‘15-minute usage’ as the proportion of consumption-related trips made within a 15-minute walk from home. Our findings show that the median resident makes only 14% of daily consumption trips locally. Differences in access to local amenities can explain 84% and 74% of the variation in 15-minute usage across and within urban areas, respectively. Historical data from New York zoning policies suggest a causal relationship between local access and 15-minute usage. However, we find a trade-off: increased local usage correlates with higher experienced segregation for low-income residents, signaling potential socio-economic challenges in achieving local living.
[CSSSA Webinar] The 15-minute City Quantified – Dr. Arianna Salazar-Miranda
This tutorial will take attendees through a quick overview of Mesa (Agent Based modelling in Python) and then go step by step through a tutorial of Mesa-Geo. For this tutorial, we will build a Geo-Schelling Model that takes the seminal Schelling Segregation Model, but this time applies to the countries of Europe. This tutorial will provide users a basic understanding of integrating GIS into Mesa and ABMs. Speakers: Tom Pike, Jacqueline Kazil, and Boyu Wang
[CSSSA Tutorials] Mesa-Geo: Agent-Based Modeling and Geographical Information System in Python
Dr. Ketika Garg presents her work on modeling information search. Abstract: Searching for resources or information often requires balancing exploration for more options and exploitation of previously found ones. Individuals searching in a group face another trade-off: whether to engage in social learning to exploit the solutions found by others or to search solitarily for unexplored solutions. Social learning can better exploit learned information and decrease the costs of finding new resources, but excessive social learning can lead to over-exploitation and too little exploration for new solutions. I will talk about a model of collective foraging designed to study these trade-offs. First, I will discuss how individual search strategies affect the costs and benefits of social learning under different problem-spaces and social factors, as well as the optimal ways by which groups can be collectively efficient in finding solutions. I will then discuss results from an evolutionary model that show how social environments can, in turn, shape the evolution of individual search strategies, and the need for systems to balance explore-exploit at both individual and collective levels for efficient search. Bio: Ketika Garg is a multidisciplinary cognitive scientist, with a background in biological sciences, and a broad interest in adaptive behavior. She studies how adaptive behavior unfolds at various scales of organization, from individuals to groups, and how they shape each other. Her work often leverages foraging models, where individuals or groups search for resources like food or information in the face of various constraints and affordances, posed by their physical and social environments, and cognitive capacities. She is currently a postdoctoral researcher at CalTech, where she is further exploring the interplay between social environments and individual decisions across varied contexts: extending from traditional foraging paradigms with ethological perspectives to the emerging, new landscapes of social media. She is from India but has lived in the US for the past few years, where she is still searching for the perfect samosas.
[CSSSA Webinar] Modeling Information Search – Dr. Ketika Garg
We are very excited to host Mirta Galesic presenting "Beyond collective intelligence: Collective adaptation." Abstract: I will describe a conceptual framework that we have recently developed for studying collective adaptation: the process of iterative co-adaptation of cognitive strategies, social environments, and problem structures. Going beyond searching for “intelligent” collectives, we integrate research from different disciplines to start understanding why similar collectives can follow very different and sometimes counterintuitive trajectories. This framework can help social scientists to begin understanding and predicting currently puzzling and often unexpected collective phenomena, from conspiracy theories and extreme polarization to difficulties in addressing climate change and pandemics. We propose different approaches to study collective adaptation, including computational models and empirical data collection. The framework of collective adaptation enables the integration and formalization of knowledge about human collective phenomena and opens doors to a more rigorous, transdisciplinary pursuit of important outstanding questions about human sociality. Paper: https://royalsocietypublishing.org/doi/10.1098/rsif.2022.0736 Bio: Mirta Galesic is Professor at the Santa Fe Institute, Resident Faculty at the Complexity Science Hub in Vienna, Austria and at the Vermont Complex Systems Center, UVM. She studies how simple cognitive mechanisms interact with social and physical environments to produce seemingly complex social phenomena. Her projects focus on developing empirically grounded computational models of social judgments, social learning, collective problem solving, and belief dynamics. She is also interested in how people understand and cope with uncertainty and complexity inherent in many everyday decisions.
[CSSSA Webinar] Beyond Collective Intelligence: Collective Adaptation – Dr. Mirta Galesic
“Pattern-oriented modeling” (POM) is a set of strategies for addressing two of the biggest challenges in agent-based modeling: deciding what should and should not be in a model, and modeling behavior. In POM, we identify a set of patterns observed in the real system and then (a) design the model so that the patterns could emerge, and (b) test alternative theories for agent behavior by how well they cause the model to reproduce the patterns. Testing and developing theory in this way is a new, exciting, and valuable approach to the science of complex systems.
[CSSSA Webinar] Pattern-oriented Modeling: Turning Simulation into Science – Dr. Steve Railsback
Social media has linked people on a global scale, transforming how we communicate and interact by sharing not only ideas, but also emotions and feelings. The massive interconnectedness created new vulnerabilities in the form of societal conflict, mistrust and deteriorating mental health. I describe the tools my group developed to recognize emotions in online discussions at scale and show how they help study collective social phenomena. One such phenomenon is affective polarization, which means that political factions not only disagree on policy issues but also dislike and distrust each other. I show that affective polarization exists in online interactions, with same-ideology users, e.g., liberals or conservatives, expressing warmer feelings toward each other than to opposite-ideology users. I also show that emotions structure social networks: interactions between users who are close to each other within the network elicit positive emotions, while more distant interactions have more anger and disgust. These findings are consistent across diverse datasets and languages, spanning discussions on topics such as the Covid-19 pandemic, abortion, and the 2017 French Election. Our research provides new insights into the complex social dynamics of collective emotions with implications for political discourse.
[CSSSA Webinar] Collective Emotions and Polarization on Social Media – Dr. Kristina Lerman
Dr. Balaraju Battu from New York University Abu Dhabi presents his research on understanding the evolutionary dynamics of cooperation and social norms. Abstract: Social dilemmas such as hunting large game, sharing meat, conserving common resources, and engaging in warfare have been critical to the survival of human communities because they provided benefits to all group members, regardless of their contributions. In these contexts, it has been observed that an individual’s cooperative action depends on the cooperative actions of their group members. However, a few members’ selfish interests can destroy conditional cooperation in their social group. The evolution of cooperation is contingent on the critical mass of individuals who adopt and reinforce the social norm, thereby triggering others to conform. The study delves into the importance of the critical mass of stubborn altruists in enforcing social norms and establishing conditional cooperation. However, very few bad apples---unconditional defectors---are sufficient to destroy conditional cooperation, even in the presence of a few good apples.
[CSSSA Webinar] A Few Bad Apples Can Destroy the Reciprocal Nature of the Majority
Abstract NetLogo, probably the most widely used agent-based modeling (ABM) platform, is designed to have a "low threshold and high ceiling," meaning that it is easy to get started but can also be used for advanced research. NetLogo has succeeded on both of these fronts, being widely used both in educational settings and by professional researchers. This session will provide a tour of some of NetLogo's advanced features which "raise the ceiling" of what it can do including: new BehaviorSpace features for running complex experiments, BehaviorSearch for finding parameter combinations in high dimensional search spaces that lead to target emergent patterns, discrete event scheduling as an alternative to tick-based scheduling, LevelsSpace for embedding ABMs inside of other ABMs, and the much-improved GIS extension. Speaker's Bio Jacob Kelter is a PhD candidate in computer science and learning sciences at Northwestern University where he is a member of the Center for Connected Learning and Computer Based modeling which develops and maintains NetLogo. His dissertation research consists of two parts: (1) prototyping Morfli, a "morphable" learning infrastructure, which is designed based on complex systems principles to foster ongoing educational improvement by creating feedback loops between students, educators, and researchers and sharing findings/tools across the educational system and (2) working with a materials science and engineering (MSE) professor to re-design the introductory MSE course around agent-based models and then studying the learning outcomes. He also has published ABM-based studies of quadratic voting and the effect of introducing land constraints to macroeconomic ABMs.
[CSSSA Tutorial] NetLogo's High Ceiling: A Tour of Advanced NetLogo Features
Load More… Subscribe

Computational Social Science (CSS) is the science that investigates social and behavioral dynamics through social simulation, social network analysis, and social media analysis. The Computational Social Science Society of the Americas (CSSSA) is a professional society that aims to advance the field of computational social science in all its areas, including basic and applied orientations, by holding conferences and workshops, promoting standards of scientific excellence in research, teaching, and publishing research findings and results.

CSSSA has the goals of:

1) Improve the scientific credibility of computational social science.

2) Maintain an institutionally neutral society characterized by scientific integrity.

3) Promote the advancement of computational social science through scientific exchange, transparency, and open discussion.

Scientific progress requires scientists to build upon previous research and developments of their peers and give explicit credit to these peers — to foster trust and integrity. Publishing your research, providing transparency into your algorithms and computer models is strongly encouraged via the Open Agent-Based Modeling Consortium. In addition to providing an open repository for models and simulations OpenABM also provides a valuable forum for collaboration, learning, and networking.

OpenABM