Developing the Model Builder Web Tool

To develop our Model Builder web tool, BioInteractive worked with Jon Darkow, an educator in Ohio who has extensive experience in developing and using models with his students. We got a chance to discuss this development process with Jon, including the Model Builder features he’s most excited to share with other educators, as part of the following video and Q&A.

 

Tell us a little about yourself, including your classroom context or anything that would be helpful for people to know about you in terms of your experience in science education.

Sure. I have been a science teacher for the past 19 years at Seneca East High School, which is a rural school in Northwest Ohio. I currently teach anatomy and physiology, AP Biology, biology, and physics.

Most importantly, I’ve been using modeling, and specifically computational modeling, with my students for about 16 years, so I’ve been trying to do modeling with my students for a long time and with varying degrees of success. My main interest with Model Builder is helping to bridge my passion for modeling and modeling with students, and the rich understanding that modeling can bring to a science classroom, to connect with a broader audience.

How would you summarize modeling, and its applicability in the science classroom, in terms of what educators need to know?

Models are representations of ideas. Models are a way to talk about ideas with people besides just using words. Models can be diagrams. They can be illustrations. But they become a useful way to communicate ideas, and for students to have something in common when you discuss systems and ask what will happen if we interfere with those systems.

Why do you find modeling as a practice so meaningful for your students?

For my students, I think modeling is a great way for a group — either a small group or a whole class — to discuss an idea and discuss complex chains of causation and interdependent causation in complex systems. I think students have a hard time thinking about that using only text. For example, what happens when the otters increase or what happens when the sea urchins increase in a kelp forest food web? It’s hard for them to see the chain of causation in a complex food web. I think visual diagrams and models become a nice way to talk about cause-and-effect relationships.

Additionally, when students build their own models — which I think is the holy grail of modeling — it really gives them insights about relationships that they had not thought about before. Models aren’t just about learning a system. Students start uncovering things they hadn’t thought about before as they are starting to design their own models.

As an example: If light causes photosynthesis, what’s really happening in between those things? And you can keep learning more and more as you start trying to make vague ideas more precise. That’s what I think is really the goal of modeling: to try to take these vague notions we have and make them much more precise. And to answer questions like what’s the direction of causality? Is a factor increasing or decreasing? How are various components all working together?

That’s a great place to segue into discussing Model Builder itself. How would you summarize Model Builder for an educator?

Model Builder is a free online tool. Its main feature is that you can create a reference model and assign a self-check version to students. Students can then link the objects in the model together. Model Builder will check the reference model against how the students are building their model and give them automated feedback. It’s a way for students to learn how scientists build these cause-and-effect models, as well as other types of models.

Which features of Model Builder will be most important for your classroom implementation?

A Model Builder feature that I really like is the self-check function for students. I use it all the time. With the self-check feature, students are already given the objects, and they’re just trying to see how they are linked together. This has become a great activity for me to intervene with students and walk around the room as students are struggling to see these cause-and-effect relationships.

What I love about the self-check is students can check their models to see if the models are right or wrong, see how many mistakes they have, etc. The automated feedback is really the novel feature of Model Builder that sets it apart from other modeling applications.

There are lots of applications where you can build models — you can build a model in Google Slides if you want. But you can’t get automated feedback with other modeling programs like you can with Model Builder. This iterative process of making connections, checking the model, then revising connections and checking again has been really neat for students to see. They can end up building a complex model just by interconnecting and checking it repeatedly.

Automated feedback allows me to go around and explain aspects to students who are really struggling, and to see, as students are struggling, which parts of the model they are struggling with. It offers an opportunity for me to see which part of a system students don’t understand, so I get insights about what to teach.

Another aspect of Model Builder I like is the creation mode, where students create models. You can build models with lots of other programs. What’s nice about Model Builder is that students can check their models, but they can also create their own using the same platform.

So, as an example, we just built food webs, and students were able to quickly make food web models. My students knew how to navigate everything, so they could build these models easily. Students built their own food web of animals they researched, and then I had them present their own models, so I could ask questions and have students answer as a group.

What is one thing that educators should know before they use Model Builder?

BioInteractive has created a large library of models for students to use over a large scope of the biology curriculum, including molecular genetics, ecology, evolution, and physiology. The BioInteractive models are in Model Builder to help you get started and understand how these models are built. There are also materials for educators on the Model Builder webpage, including pedagogy and user guides and a catalog with “answer keys” on how to build all the models.

Which feature would you say has the biggest learning curve in using Model Builder?

The Grading Tool has the largest learning curve. When you create a model with Model Builder, you will end up saving the model as a ZIP file that will generate a self-check model, an assessment model, and the reference model. Then you can share the assessment model with students. Unlike the self-check feature, with the assessment model, students will be given the objects, and they’ll link those objects together in a particular way but aren’t given automated feedback. They then save that model and share it with you.

You can use the Grading Tool to check students’ models, and all the students’ assessment models will be auto-graded against the reference model. The Grading Tool requires a few more steps that are a bit more complicated than just building a model or using one of these self-check models.

What is something that you have learned about yourself as an educator or your classroom practice through the experience of helping to build Model Builder? And how has your thinking about modeling changed over this process?

It’s been a fantastic experience working with BioInteractive on this project. I feel very fortunate to be able to do that. It has let me understand how to make things simpler for students. As I have tried to make models and modeling more appropriate for a larger audience, it’s been better for my own students too. There’s always a knowledge gap between you and your students. Working on Model Builder helped me to clarify the steps that I’ve tried to do — going from students not knowing anything about modeling to trying to build computational models. This has been a great experience to try to bridge that gap and help them understand how models represent phenomena in science and how we can evaluate our models.

Another powerful aspect of Model Builder is that the tool can be used throughout your curriculum. Students start seeing connections across the curriculum, and they get a lot better at building models as a practice.

What information about Model Builder or modeling would you like people to know about that we haven’t discussed?

In terms of pedagogical approaches, there are several ways you can use modeling in the classroom. For one, you can give students a text, video, or simulation and then have them build a model that is represented by that resource. This approach has students re-represent the content to see if they’re understanding cause-and-effect relationships.

Additionally, Model Builder has positive and negative connectors, and those help to build students’ mathematical and quantitative literacy. That I think is an important tool, without having to involve any numbers or algebra. Model Builder is visual, so it helps students think quantitatively without thinking specifically about math.

With complicated models, I will partially build the self-check model, save the model, and share with students. Then students can interact with and complete a complicated model that has been mostly built for them. Sharing partially built models scaffolds learning more complicated models and allows students to still successfully make connections.