How to Validate Your Model Empirically (Part 1)
Tips on testing model results and mapping models to real data.
Qiankun Zhong is a postdoc researcher at Center for Machines and Humans, Max Planck Institute for Human Development. Her research interests include cultural evolution, organizational communication, complex system, and collective actions. She uses cultural evolution theories to understand how the technological environment shapes human collective intelligence and cooperative behavior, especially in the current AI environment.
I am a post-doctoral researcher at the Center for Humans and Machines, Max Planck Institute for Human Development. I received my PhD in Communication from University of California, Davis, with a Designated Emphasis in Computational Social Science in 2023. I am interested in the interaction between institutions and culture in influencing cooperative behaviors and democratic processes, especially in the current media and AI environment.
Through my academic journey, I have focused on one key question: How does technology shape human societies’ evolutionary process? My research tries to answer this in three ways. First, I test whether the evolutionary mechanisms that explain human collective behaviors and norms throughout history still apply to the current fast-changing technological environment and AI systems. Second, I explore how modern technology, especially generative AI, could shift social learning heuristics, leading to different collective outcomes. Third, I use evolutionary theory and computational methods to reconcile conflicts between social theories, with special focus on cultural sociology and (neo)institutional studies.
My research is rooted in the traditions of communication and cultural evolution and tries to answer new questions that emerge from the current technological environment. My research appeared in journals and conference proceedings from various disciplines, including Scientific Reports, ACM CSCW, AAAI ICWSM, NPJ Complexity, Social Science Computer Review, and Entropy.
Tips on testing model results and mapping models to real data.
Notes on how to learn modeling by reading broadly and thinking about process.
How information control can trade off robustness and panic in crises.
When does an agent understand the logic of a game?