Interesting question - I'm trying to think of how to measure that.

I have a question: how much access do you have to resources (specifically, human subjects)? If you can get access to, say, twenty or more people, then you could randomly assign ten of them to run without footware and ten to run with, and measure their performance, such as how long they took to run a certain track distance). You could then run a statistical test to determine whether the runners with shoes or the runners without were faster. You need the statistical test to determine whether the faster group was faster because of your treatment or because of the random selection of better athletes in that group. If you need help with the statistics later, I would be more than happy to help you with it.

If you don't have access to that many test subjects, you could perform multiple trial runs with yourself, on the track, with and without shoes. You would have to do so on different days, and once again you would need to randomize the order in which you do your runs, because that will affect the data (after your first few, your performance will drop because you're tired, but later on, your performance will improve because you'll be in better shape, regardless of your footwear). Of course, this also would only be measuring the effects of footwear on running speed for YOU - somebody else might do better with a different kind of (or lack of) shoe.

Your project sounds fun and interesting.I wonder what sort of data you can get with a bathroom scale and, if you have it, something like a wii fitness program with the thing you stand on that measures forces and balance.

Electronic bathroom scales can be a real problem - I got a $20 one at Walgreens that basically gave out fairly random weights - it had force* sensors on the 4 corners and seemed to just record whatever weight appeared after a set period of time of around 3-4 sec, even if the numbers were jumping around. It was fun to take apart.

My better electronic scale keeps showing weights for as long as I move around on it and stops only when I stand still for 3-4 sec. But I don't see any interesting differences among the weight fluctuations as I move around on it in different shows [see below].

An 'old fashioned' bathroom scale with a dial would register whatever forces* it was sensing at the moment, so you could have an observer who would try to see what the force* was as you ran across the scale in different shoes. But it might be hard to determine what number the scale reached when someone just stepped on it in the process of running.

My main contribution is 2 unusual kinds of shoes you might not have thought of:Z-coils, with big springs in the heels andMBT's [Masai Barefoot Technology] - the 'rocking' shoes.

But I guess the challenge would be to find those shoes, and you might not choose to 'advertise' those types of shoes if you don't have any data on them.

There's probably a lot of interesting research on the web about shoe design vs performance; some of it will be doubtful, because it will just be uncontrolled research by the shoe company to shoe how good its shoes are supposed to be.

Don't be reluctant to present negative results, such as: Our measurements didn't show any reproducible difference between the different shoes we tested. That's a common research result that doesn't get published very much. My son was trying to test color preference in his cockatiel for a science project, and he had the cockatiel choose which colored paper clip to play with. The cockatiel basically liked playing with the paper clips and didn't seem to care what color it was. So maybe the cockatiel didn't have any color preference, or maybe the experimental design just wasn't good for showing what the color preference was. He could have done the experiment in lots of different ways.