Computer Science in Science PD: Using Computer Models in Science - Discussion

Typically, there is always human error when completing a lab in the classroom.

It is similar because it allows the students to experience their learning in an analytic way, but the traditional lab setting allows for more tactile inclusion of experiences.

One of the biggest differences is that Labbs allow for more human error. It is good idea to have the students thinking in a more analytic way that allows them to not be to concerned of making mistakes and/or being limited to just a few trials.

I agree 100% that time can become a huge burden. This allows us too cut back on some of the barriers that keep the students from getting discouraged.

The setup with a computer simulation as compared to a traditional experimental design is completely different. Instead of physical instruments such as beakers and lab coats we will be running simulations. That being said we will not be limited to one controlled experiment.

I couldn’t agree more. The ability to focus on the data will really enhance the discussion and help students understand it on a deeper level.

Using computational models instead of wet labs is a wonderful way of splitting a large group into two. Teaching 30+ students and wanting to do multiple labs is great, but funding and time get in the way and mostly just plain time is the biggest issue. Having 30+ students and only 1 teacher is really the big issue with safety and squirrely 8th graders. If you have computers, you can split the group into two and have half doing computer models and the other half doing a wet lab on the same thing. It would be a great way to show human error and all of the “control” variables effects as well as see how well the computational model actually works. It also keeps the safety issues to a minimum as the kiddos working on the computers are all seated and working quietly and the teacher can be available for the wet lab mishaps that invariably happen.

Experimental design is similar in both a traditional lab setting and with computational models. Computational models could bring the following differences:
-ease of repetition of trials
-lower cost in terms of supplies
-time savings
-quick generation of data
-simulations or parts of simulations could be used when communicating results

Don’t mistakes like this make a lasting impression? I agree that eliminating some of these errors keeps everyone on the same page, but an awareness of the importance of carefully following instructions would be lost. Or maybe not…they could run the wrong simulation or something. :smile:

One of the major limitations in science class is time. Using a computation model students can repeat an experiment or make changes to the design to see different results.

Even though I don’t teach science I still tie the idea of making predictions and forming hypothesis in my class. I give students a problem and I have them predict what will happen. One of the aspects of using computer modeling that I really like is we can create simulated models to help students visualize results.

When experimenting you must not only have a hypothesis, but you must gather evidence for argumentation and analysis. I think that developing models that you can experiment wit through coding is a great way to do this.

Computer modeling is different from the labs I do in my classroom because it is easier to do multiple trials, change behaviors, there is no human error, and hypothesis can be built.

In the classroom, labs are usually deterministic – looking for one outcome. It’s difficult to bring in multiple variables because of time, money, supplies/equipment on hand. You’re usually looking for students to come out of a lab with a specific concept. You, also, have a lot of human error since students are just learning new lab skills and may not be adept at using new equipment or methods.

Using computational models is more realistic in that you don’t get the human error and you can program your model to include whatever factors you want to control. You, also, have the luxury of saving time with a model and so, can run as many trials as you want, can run many different variables, as well as run a simulation that has multiple variables to have students see what happens in real world situations.

I completely agree! This opens up many possibilities since the students would be able to run many experiments in our short class periods and would have access to experiments that were previously unavailable to them because of lack of materials.

I echo what others have said earlier about computational models making a lab flexible. They allow a work-around for the (logistics, time, resources, etc) restrictions in our traditional classrooms. The use of computational models into curriculum isn’t much different from experimental design; rather, it makes it easier to due in a classroom since students can revise and rerun experiments more times conducting them physically.

Our school is an International Baccalaureate school so we are constantly asking our kids to do some sort of a project to demonstrate their learning. As such, they have to follow steps that parallel the scientific method.
With computational models, they could actually see these projects through to a new level and develop a deeper understanding. Instead of just making a prediction based on research from the internet or books, they could run these models to get a better representation of possible outcomes for their projects.

This is so true! So often the kids only test a hypothesis a few times because either it’s too expensive or too time consuming for them. Using the computational modeling would allow them greater data collection and better overall results.

I’m not sure I agree. Shouldn’t students set their model and still hypothesize as to the outcome? In my mind, this would ensure that they are thinking of how the variables will affect one another, rather than randomly assigning properties.

As I said in the last forum, one of the main benefits of a computational model is its cleanliness, both in the actual realm (nothing to clean up) and, more importantly, in the data that are generated. There is little chance for human error, and students can clearkly see how changing a variable will affect an outcome.