We tried our best in controlling as many variables as possible, and, for the variables which we couldn’t control, we had enough variety in those variables to disregard them as not statistically significant.

 

Firstly, anything that could affect the readings on the multimeter was accounted for. We measured the resistance in the alligator clips and were delighted to discover that it was so small that the multimeter literally displayed zero ohms.

 

We need our simulation of rainy conditions to be as accurate as possible to actual rain coming down the downspout. We were unable to connect our generator directly to Jason's drainage system without ruining his drainage system, so we bought our own piece of tin downspout. The downspout piece was only two meters long, however, so we could not accurately simulate the 10 meters of gravitational acceleration that the water would be subjected to with a two meter pipe.

 

Luckily, the hose, at maximum power, shoots water out at 7m/s. That plus the length of the downspout roughly accounted for the amount of gravitational acceleration the water will be subjected to. Since any small amount of extra or missing Joules of energy the water carries are probably going to be lost through friction from going down the gutter and turning the wheel, we aren’t too concerned about this aspect.

 

We also tried to control for the mass of the water by first testing the flow rate of the hose before inserting it into our downspout. Most of the time the flow rate of the hose was consistent but once in awhile the flow rate went way down and we had to shut down the hose, then spending the next 10 minutes walking around the yard looking for obstructions.

 

By keeping constant the mass and the acceleration of the water, we effectively controlled for the amount of force that the water will have when it strikes our water wheel, but despite this careful controlling of the two most critical variables, some variables still slipped through our scrutiny and we were unable to account for them.