One of my favorite labs to emphasize testing and controlling variables is the classic pendulum experiment. Typically this is framed in a procedure prompting students to perform trials where they change one variable about the pendulum set up and collect data about how the frequency of the swing is affected. This is repeated for the mass of the bob, the string length, and the pull back amplitude. It's a surprising conclusion for many that only the length of the string affects the frequency but the numbers don't lie.
Overall, we probably spent close to 4-5 partial days working on the prototypes before they were asked to demonstrate their effectiveness and accomplishing the task. In this time, they needed to figure out away to collect enough data to model their system (either graphically or mathematically) so that they would have something to guide their design during the demonstration.
Overall I was very happy with the results of this project. I felt that it truly challenged the thinking of many of my students and it led to some really interesting conversations about the hows and whys of setting up experiments within different groups.
I also asked for feedback from students about what they thought about the project itself. The results were split many students liked the challenge and open-endedness of the project but that also caused a great deal of anxiety for some. There is also a disconnect where students still don't see projects as a way of learning. I hope to be more explicit with the learning targets that I hope to hit in the next project.
Father, Physics Teacher, Knowles Fellow, Friend, Techie, and Musician