ASB 591 / CSE 591: Computer Simulation in the Social Sciences

Spring 2006

Professor: Marco Janssen

7 groups of students worked on projects where they developed a simulation model in Netlogo of social phenomena of their interests.

An Open House to see the projects is organized at May 9, from 2pm to 4pm, in Matthews Hall 103-112

 

Modeling the Adaptive Behavior of Leadership Strategies in Middle-Range Societies

J Benton (Computer Science) and M. Scott Thompson (Anthropology) developed an agent based model to analyze variability in leadership strategies using Blanton et al.’s (1996) corporate-network axis. Blanton et al.’s (1996) “Dual Processual Theory” proposes that leadership strategies vary along a continuous axis, with the “corporate” leadership strategy on one pole and the “network” leadership strategy on the opposite pole. In the model, a leader’s behavior is governed by a weighted parameter.  Corporate behavior emphasizes a leader’s connections to households, while network behavior favors the acquisition of goods. We evaluate a hypothesis that change in the social perception of a leader’s responsibility is a principle factor in change in leadership strategies.  [Black, J. A., R. L. Oliver, J. P. Howell, and J. P. King, 2006. A Dynamic System Simulation of Leader and Group Effects on Context for Learning. Leadership Quarterly 17: 39 – 56.]. 

PALM: Prehistoric Agricultural Landuse Model

 

Tanjot Bhatia (Computer Science) and Matt Peeples (Anthropology) developed an agent-based model on movement of households in the prehistoric SouthWest of the USA. Households of different sizes perform agricultural activities, which may deplete the resource, and may move when they see better opportunities to feed their family. They may move within their settlement, or between settlements (but then they can not take their stored food items with them). Besides resources, the social capital (accumulated occupation) is a factor that attracts people to a settlement. The more social capital is important in making decisions where to settle, the more the agents aggregate in one place, and societies experience massive movements from time to time.

 

 

 

 

Tuberculosis with migration

Amy Farnbach (Anthropology) and Mike Rush (Computer Science) developed a system dynamics model of the impact of migration to tuberculosis occurrence in historic America. The disease needs to have a minimum population level to sustain. Amy and Mike show that small populations are sensitive to migration of agents through the settlement. An agent-based model version was also developed, but there is not enough information on the mechanisms of contract transmission of TB to reproduce observed data.

 

 

 

 

 

Reputation of Place at its role for agent aggregation 

Chris Watkins (Anthropology) and Jin Zhou (Computer Science) developed an agent-based model to study the emergence of market places. Agents move around on the landscape and try to exchange goods. When they have a successful transaction, that place gets a positive reputation. Agents are more likely to move to places with positive reputation. Depending on the decay rate of reputation, the speed of the spread of reputation, and the risk attitude of agents, places emerge where agents will come to exchange goods.

 

 

 

 

 

 

FLOSS Sim: Simulating the Evolution of Open Source Software Projects

John Hooper (Anthropology), Olga Kornienko (Family Studies) and Nicholas Radtke (Computer Science) developed and agent-based model to study the open source software development and use community. Agents can consumer and contribute to a set of different projects. Agents have different needs and when they contribute to the development of a project, the characteristics of a project change too (towards needs of developers). Agents can consume projects, downloading, based on expected utility of using the software. The more is contributed to a project, the project reaches different stages of development. The model is generating simulated data that can be compared against observations (distribution of persons per project, projects per person, stages in which projects remain, etc.).

 

 

 

 

 

 

Agent-based Pottery Exchange and Distribution

Yu-Ru Lin (Computer Science) and Joshua Watts (Anthropology) developed an agent-based model to explain patterns of pottery exchange in the prehistoric Southwest of the USA. Three types of agents are considered: farmers, potters and decorative potters, who exchange goods to meet their needs. When agents need to buy or sell, they need to find the right partner to exchange say food for pottery. Different rules for finding a partner are explored (random (as long as match is made), restriction on kinship, and restriction on distance). The model is doing a reasonable job to generate patterns of pottery styles that are observed in archaeological data.

 

 

 

 

Modeling Resource Levels and Population Levels on a Hypothetical Landscape Using Migration and Sunk Costs

Jason Sperinck (Anthropology) and Pia Tripathi (Computer Science) developed an agent based model of settlements where the population may migrate to other settlements when resources get overused. They used the Janssen and Scheffer (2004) Ecology and Society papers as their starting point and included migration rules. Population may migrate to those settlements based on potentially different rules such as based on resource availability or sunk costs build up in other locations.

 

This site was last updated 05/10/06