Statistics in a Pacing Experiment

I am setting up an experiment and I need help with the design because I would like to run different stats tests on the results. I am testing to find which of three pacing methods is most effective for runners: a lighting system, a GPS watch, and personal feel. For each method I am asking the participant to run a the same prescribed pace for 20 meters while I use video to determine how close they are to the actual pace. The order in which the three pacing methods is also being randomized How many people should I test? Should each person be tested more than once, if so should each trial be randomized for each person? Is there a way to calculate the number of people I should test?


No cake for spunky
You should look at the design of experiment literature particularly in that inside ANOVA to begin with. This addresses many of your questions. I am not enough of an expert to comment with certainty.

You should do a power calculation using the effect and uncertainty level you feel comfortable with (that tells you how many cases you need).