Conducting Your Own Research Project

Your method of analysis greatly influences how you will conduct your experiment. One example of this occurred when our team determined how they would conduct the ball/court interaction experiments.

We had a few different ideas that we evaluated. One idea was to build a billboard that had a grid on it. The grid would be sort of a checkerboard or possibly have radial lines on it. The experiment involved videotaping balls bouncing in front of this board. The analysis of this experiment set-up would involve reviewing the tapes and determining the angle before and after the bounce, the velocity, the spin and the rebound height frame, by frame by frame by frame with the naked eye.

This seemed like a very manually intensive method of analysis to the principal investigator, Jani Pallis. Aside from the logistics of building the billboard, the cost and transporting it across the United States from California (where the team resides) to Florida (where the USTA training center is), there were some concerns about how accurate this visual method would be.

Jani knew that the team could write software to generate a grid and superimpose it over the videotape on a computer. Then the billboard "grid" would not be needed. Nasif Iskander, our co-investigator and physics teacher, had used a software package in his classroom that used each "pixel" (each dot on your computer screen) as a grid point! We would need some good reference points out on the court to make this experiment work with this method of analysis.

We chose 12 inch blocks of wood. Our software would tell us how many pixels (grid pints) this block had. This gave us a ratio of inches to pixels. We were able to use this scale to help analyze the data.

Jani later did something similar when she analyzed Pete Sampras racket head speed. John had given her the dimensions of Sampras' racket and she was able to use that as a scale in her analysis.