Science Fair Project Ideas

A simple way to conduct tests without mich equipment is to either use percieved effort or time to exaustion.

A project might be to look at how specific stimulus before a 800m time trial effects percieved effort. (Does someone watching 30 min of gladiator prior result in a different percieved effort during a TT versus 30 min of SAT questions?).

Another option is to use a compter and camera and do something about the ever popular foot strike or stride length. Does foot strike chance with increased speed? Does leg length or runner height effect this? Kinovea is a free software that lets you measure stuff from a video after you set a reference frame. Maybe ask your coach to let you film some warmup laps during practice?

No matter what you do, the data analysis is gonna be the most important part. You are gonna need at least 10 runners, otherwise it’s really not enough to be more than an anicdote.

I’ve done this stuff for classes before. It’s fun, but it’s not easy. The fact that it deal with something you like (running) is helpful.

Compare split shorts to half tights in a wind tunnel. Calculate difference at common race distances and speeds.

Do a experiment to see if water curves or does it find its level?

You could try an experiment related to fueling or hydrating before a timed run (e.g., a mile). You want to have a "control" runner with a standard/recommended pre-race routine (drink until thirsty, don't eat 3 hours before), and compare that with someone who over-hydrated or drank a Big-Gulp of Coke or Gatorade. You could also try eating a Big Mac or 5-hour Energy or a Snickers bar, etc. There have been some studies that show a performance benefit of caffeine before a race, so you could try to prove or disprove those studies.

It does help to have a group of runners to get some statistics.

Guy who runns gave some good ideas, but I think a slightly better idea would be to video runners on a treadmill and measure three things: 1. stride length, 2. cadence, 3. foot strike (fore foot, mid foot or heel strike).

you measure these three at different running speeds.

then you use your data to answer the following questions:

1. how does cadence change as speed increases?

2. how does foot strike change as speed increases?

3. how does stride length change as speed increases?

it is already well understood that speed = (stride length x cadence), your study can develop an algorithm for determining the speeds at which an athlete gets faster by increasing her stride length, versus the speeds at which she gets faster by increasing her cadence.

you can, essentially, make this as simple or complex as you feel you can handle and have the time and skills for.

cheers.

Cottonshirt wrote:

Guy who runns gave some good ideas, but I think a slightly better idea would be to video runners on a treadmill and measure three things: 1. stride length, 2. cadence, 3. foot strike (fore foot, mid foot or heel strike).

you measure these three at different running speeds.

then you use your data to answer the following questions:

1. how does cadence change as speed increases?

2. how does foot strike change as speed increases?

3. how does stride length change as speed increases?

it is already well understood that speed = (stride length x cadence), your study can develop an algorithm for determining the speeds at which an athlete gets faster by increasing her stride length, versus the speeds at which she gets faster by increasing her cadence.

you can, essentially, make this as simple or complex as you feel you can handle and have the time and skills for.

cheers.

I think it could be cool if you looked at strava data on these types of things. You could get the whole distribution of cadence and stride length and so on, and the data has already been collected, you just need a way to analyze it.

Strava is a cool idea, but not very accurate. One of the rules in scientific research is to use consistant measuring apparatus. Eg: don’t measure some stuff with calipers and then compare that with measuring tape measures stuff.

The problem with strava is that you are relying on other peoples GPS watches and the data can be inconsitant. Some garmins read higher cadence than suunto. How do you figure out stride length without measuring it? What about the effect of an uphill grade on the data? Or someone who was just feeling “off” during a run? Using past strava data means you can’t rely on your own data. Too many variables.

Cotton shirt had a really good suggestion for a potential project.