"Disinformation Spread" is a suite of computational models designed to encourage reflection on
the factors that affect belief in and the spread of disinformation.
Disinformation Spread is part of a project that I lead at the Transformative
Learning Technologies Lab based at Columbia University. The models are developed using
the NetLogo language and extend two other models -- "Spread of
Disease" (Rand &, Wilensky, 2008) and "Virus on a Network" (Stonedahl & Wilensky, 2008). Each of
those models addresses a learning need identified in previous academic research on mis- and
disinformation.
My role in this project: I conceived, designed and developed the models by remixing existing ones. In Fall 2022, I recruited a Master's student to help me turning the prototype into a research project. Since then, we have collected data through a study in which we interviewed learners testing the models and giving feedback on the learning possibilities using the models, as well as their user interface and experience. We used screen recording and task-based interviews to collect data, then analyzed data using thematic analysis and UX research techniques.
The design process involved user interviews and learning about agent-based computational modeling , more specifically using Java-based NetLogo. Below are some images of the models in their current stage of development.
image: model #1, containing components that are common to all others
image: model #2, with "instant messaging" as form of spread
image: model #3, with "social media" as form of spread
image: model #4, which adds "fact-checking" as a factor reducing
the pace of disinformation spread
My role in this project: I conceived, designed and developed the models by remixing existing ones. In Fall 2022, I recruited a Master's student to help me turning the prototype into a research project. Since then, we have collected data through a study in which we interviewed learners testing the models and giving feedback on the learning possibilities using the models, as well as their user interface and experience. We used screen recording and task-based interviews to collect data, then analyzed data using thematic analysis and UX research techniques.
The design process involved user interviews and learning about agent-based computational modeling , more specifically using Java-based NetLogo. Below are some images of the models in their current stage of development.
image: model #1, containing components that are common to all others
Click here to view the prototype. Here and here are conference papers associated with this project.