Sunday, February 28, 2016

Van Lehn Response Bottorff

When modeling Zika, students would benefit from various ideas from Van Lehn. For example, Van Lehn discusses agent-based modeling and its benefits in model construction. Besides allowing the construction of an incredibly diverse set of models, agent-based modeling allows computers to execute the model. However, students may need additional resources to learn formal language, i.e. not drawings or natural language. There are undoubtedly many misconceptions for students learning formal language that need to be addressed.

Zika is easily modeled using agents and spurs practice using formal language. Students can quickly investigate complicated manipulations of blocks even in StarLogo to model advanced aspects of Zika.

He shows that model creation is more effective than using previously made model. Students therefore are allowed sufficient time and space to think and revise. It is especially important to emphasize model creation when there is no standard model for use in new cases such as Zika. Developing the skills to make new and creative models is a necessary skill. He assesses models by both the product and the process of creation, emphasizing the importance of both.

Another benefit seen of modeling by Van Lehn is the improvement of domain knowledge alongside model creation and use. Using models can remove misconceptions, make existing knowledge more relevant and useful, and simply add new domain knowledge. Developing additional domain knowledge through and alongside the use of models can allow programmers to create models that more accurately represent nature, even Zika.


  1. I think its important to realize the possible hindrances students might face when trying to construct models, especially for a system like Zika when we do not know everything we can about it. I agree that some students will most likely need help with formal language, and while VanLehn dedicates a decent part of his paper to "difficulties" most of these revolve around changing aspect of the computer program itself. Are there any ways that difficulties/misconceptions can be overcome outside of the computer based modeling program?
    I'm not sure if we have the full answer yet, but what could this look like in our classrooms and how could we make sure we know how to troubleshoot in case our students are having difficulties?

  2. You make a good point about there not being a standard model of Zika and, thus, no pre-made model for students to use for comparisons. I wonder if it would be useful to have students first model another virus, compare their models to known standards, revise their constructed model, and THEN have students attempt to model Zika. This process could help students analyze different facets of disease transmission and determine what some "standard" disease models look like.