NetLogo User Community Models

(back to the NetLogo User Community Models)

Download If clicking does not initiate a download, try right clicking or control clicking and choosing "Save" or "Download".(The run link is disabled because this model uses external files.)

WHAT IS IT?

This model is an agent-based model that simulates member motivation and contribution in a conversation-based online community such as UseNet groups or web forums. Individual agents imitate members in an online community and can make decisions to read and post messages based on perceived benefits and costs of doing so. It can be used to run virtual experiments to develop theories about individual behaviors and social dynamics in online communities and to inform the design of real-life online communities.

HOW IT WORKS

Each tick simulates a day in real life. At each tick, community members decide whether to participate by reading messages and whether to contribute by posting messages. They consider four types of benefits from receiving information, sharing information, social connections with the community and its members, fun and reputation. They also incur costs in reading and posting messages.

A member reads and/or posts messages when (expected) benefits from reading and/or posting messages exceed costs of reading and/or posting messages. Members who remain inactive for a long period of time are regarded as having left the community.

HOW TO USE IT

Use the sliders initial-member, initial-message, and run-time to specify the initial condition of the community. Use the chooser comm-type to specify the type of the community: technical, interest, support, and political. Members of different types of communities diff in their motivations. For instance, members of technical communities such as a JAVA developer group care more about finding information about JAVA development than making friends whereas members of a health support group care about both finding information and giving and receiving social support.

Click the SETUP button to create a virtual community. Click GO to run the simulation.

THINGS TO NOTICE

- Watch how members enter and exit the community.
- Watch what happens in the Member Statistics Plot.
- Watch what happens in the Member Entry and Exit Plot.

The program also outputs statistics to a text file for further analysis. For example, you can track member composition in terms of history and interests, member entry and exit and visit frequency, various type of benefits individual members experience over time, or the number of messages posted by members over time.

THINGS TO TRY

- Try to change the number of initial members and see what happens.
- Try to change the type of community and see what happens.

Or you can use the Behavior Space under "Tools" to vary pany parameters in the model to run virtual experiments. We have included the experiment we ran for the Human-Computer Interaction paper as an example.

Vary variables as follows:
["initial-member" 30]
["initial-message" 30]
["run-time" 250]
["comm-type" "interest"]
["moderation" "no" "yes"]
["personalize" "no" "yes"]
["topic-breadth" 2 6 10]
["offtopic-interest" 0.1 0.3 0.5]
["filter-precision" 1 0.8 0.6]

Repetitions: 5

Measure runs using these reporters:
run-number
count subjects
count subjects with [membership = 1]
count subjects with [message-self > 0]
count messages
count messages with [thread = -1]

EXTENDING THE MODEL

The model can be extended in many ways. It can be extended to simulate other types of online communities. Currently, it is built to simulate conversation-based communities and you can add new actions to simulate production-based communities like Wikipedia. That requies the creation of new objects such as articles or user profiles and the relationships between members and these objects.

It can be extended to incorporate more social science theories or features of online communities. Currently, we draw on a selective set of theories such as value expectancy theory, Collective Effort Models, information overload, group identity and interpersonal bonds, reciprocity, etc. Insights from other theories such as public goods, critical mass, task interdependence, network externality and evolution can be incorporated to enrich the model. Or it can be extended by simulating the working of new features such as leader board, subgroups, user ratings of messages, etc.

CREDITS AND REFERENCES

This model is authoried by Yuqing Ren at Univeristy of Minnesota and Robert E. Kraut at Carnegie Mellon University. You can find a detailed description in our paper forthcoming at Human-Computer Interaction.

To refer to this model in academic publications, please use:

Ren, Y. & Kraut, R. forthcoming. Agent-Based Modeling to Inform Online Community Design: Impact of Topical Breadth, Message Volume, and Discussion Moderation on Member Commitment and Contribution. Human-Computer Interaction.