Conducting an experiment is the same in any discipline of science; that is why the scientific method is so critical. It gives scientists a path to the most accurate data.However, computational experiments allow scientists (and students!) to explore the effect of variables in a time and cost sensitive way. This increases access to inquiry based learning for students without the time or resources to pursue a longer or more elaborate traditional experimental model.
Good point! I was always so frustrated on lab day when the materials didn't stick to script! CS models help to illustrate the concept without having the interruption of hidden curriculum, or user error, on arcane materials.
computational models allow for repeated trials, middle school students can figure out, by trial and error, if they have a solid control set up. The computational modeling process is the very similar to experimental design process - they can work in harmony to build content knowledge and skills.
Similarities include the need for controls and multiple trials. Differences include the ease with which trials can be repeated.
In class, students are given an experiment, create a hypothesis, and then test it. But with computational models, they are able to create their own parameters, run the experiment, and then create a hypothesis from the information they receive.
Conducting experiments in class requires the availability of a wide variety of material, a careful set-up and execution. Mistakes can cause real damage. Repeating the experiment would require repeated set-up and additional material. It would also take time and expense. Computational models can be built and reconfigured quickly, at no cost and can be adjusted to manipulate variables easily.
I always to talk with my classes about the importance of completing an experiement several times to check for accuracy etc.. but this is often not possible in a classroom setting. Using CS will allow us to ensure we are being accurate and easily make changes where needed.
Using computer models allows time for multiple trials which might not be doable for the hands on labs due to time constraints and the cost of the consumables involved. With the computer models there is also more flexibilty in testing different variables. In many instances the computer model would be easier to design a truly controlled experiment.
Experiments in class are difficult because of buying the supplies and the time factor... class periods are 55 minutes. With computer models, the time factor is not a problem. I also like the fact that you can have different outputs of a set of inputs and students can see randomness in those outputs. More like the real world.
The physical aspect of students participation in unplugged activities can be enhanced by the computational experiments and vice versa if we are thoughtful in our choice of activities and models.
Dissections are one of the most memorable labs we do in 7th grade at my school. Performing this lab physically for students has many advantages, but also many disadvantages. If students were to perform this lab using a computational model there would be less room for error. For example, some students cut too deep, or make cuts on the wrong part of the animal. Also, students who have certain views on animal cruelty would appreciate a computational model.
Great point! There never seems to be enough time to set up, perform, and clean up for labs!
I am not a science teacher. I am a technology teacher, but I do have some experimental aspects to my class. Usually students try out a sensor to see how it works and learn how to program it by using the sensor and observing how it works and then making a conclusion. In the experimental part of the lesson, there usually is only one or two answers without any randomness.
using computer models in science makes it easier and faster for students to see the results. Mistakes could be fixed easily and different variables could be tested quickly. This is not necessarily achievable when conducting science experiments in class. Human error and environmental effects can not be eliminated while doing hands on experiments in class.
They both have independent and dependent variables and constants. One thing is done (either virutally or in real world) and the impact on the dependent variable is observed and measured. The computer model can be easily replicated and can conduct experiments which may take a long time (plate movement) or may be too dangerous to conduct (the effect of cyanide on living human tissue). In some ways, the computer model takes the frustration and joy out of conducting real world lab experiments.
Two things I emphasize, finding that neither is intuitive to students, are the importance of controlled experiments and the need for repeated trials. It sounds like those are also features of the computer simulations.
I think what you said about taking the frustration and joy out of experiments is a really good point. I know students enjoy using computers but they seem to enjoy even more hands on experiments. So finding a good balance will be important.
I love the idea that my students may model something even though they don't have access to the materials.
Class labs and computer models are similar in that they both allow you to answer a questions through inquiry by collecting and analyzing data. They differ in that computer models allow the testing of multiple scenarios.
When I teach my students how to conduct an experiment I always focus on the independent, dependent, and controlled variables at the beginning, because my students struggle to understand how important it is to keep all trials the same. I think using computational models probably helps with keeping data consistent throughout the experimentation process.