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

Material costs are significantly reduced yet still allowing for inquiry, probably more inquiry than “live” labs.

The entire time I was watching the video, I was thinking about a lab in which there were many variables that were beyond our control that effected the outcome of our results. One variable we couldn’t control was the air conditioner. I can see how the experimental design process would allow for more accurate data and also ease of repetition.

In conducting science experiments, I would agree that time is a huge factor. When comparing the heat capacity of soil versus water, inevitably there is always several groups where the data they collect does not match their classmates data. There are a variety of reasons for this. By using the experimental design process, students could run the experiment several time and, compare the results to the other groups results. A class discussion could then take place on what variables were not taken into consideration which lead to the unexpected results.

A great advantage of using computational science in experimental design is the ability to run multiple trials quickly. Even with block scheduling, some experiments feel rushed, thereby limiting the reliability of the data. CS eliminates some of the human error associated with teens and preteens setting up science experiments.

This is a great point. I spend lots of money each year purchasing items for labs. Just swapping out 3-4 of those with a CS equivalent would be a great savings.

While both are effective teaching practices, computational models often allow for more flexibility and are easier to modify. When students are curious and begin to create their own questions, the computational models would allow them to alter the situation to explore other possibilities.

Both are effective for teaching science as they teach students to predict, observe, collect data and create hypotheses. However, computer models allow students to test things that they cannot see, they allow students to explore new possibilities and to focus on specific variables.

Computer models will help the students see what they are trying to learn, I can envision them using the program to see force and motion creating a rollercoaster, which will excite many of the students.

The ability to create several possible outcomes by manipulating variables and showing the different results is very powerful and something that cannot necessarily be replicated using something other than the use of computer models.

Usually lack of time is a big factor when completing experiments. Students create their hypotheses prior to running the experiment. By using computational models, students are able to create a design, run it, repeat, and then create their hypothesis based on the simulations.

The main differences which would make it beneficial to use computational models in my class would be the ability to perform expensive, lengthy, or experiments where the variables are too large or too small. Similarities would include using sound scientific principles to set up the experiments.

The variables in computational experiments have more flexibility than in traditional experimentation. While a traditional model does allow students to “get their hands dirty” it can be cumbersome & expensive to repeat the experiment multiple times in order to test multiple variables. Computational experimentation allows greater flexibility and repeatability.

With the computational model, you can do a much larger sampling and for a longer time than you can with traditional experiments.

The labs I conduct in class often require extensive set up and break down time. There are also some factors, such as equipment and supplies that limit the amount of labs or presentations we do. Therefore the use of a computational model provides the students with a great visual for the concept.

Excellent point! I have that same problem in my classroom.

I would love to be able to give my students the possibility to run multiple trials of an experiment. This is an important part of experimentation that I usually have to just remind students … “and of course ideally we would run multiple trials.” This would allow students to actually conduct multiple trials.

This is especially true when time is limited and students do not really get to run the experiment with some practice

Everyone seems to be in agreement that both time and money(materials) are going to be saved.

As I do not teach science classes, I do not explicitly conduct experiments in class.

Time is a limitation with many experiments, sometimes students come up with variables that they want to try out and can’t be done due to time constrains. With computer modeling they can try out their variables several times and change the variables one at a time to “see” the trend.