Well, the day finally came! My Kindergartener came home with her first Science Project assignment. As an engineer (and a Dad) I was thrilled. After discussing experiment options, my daughter decided her project would be "What makes a locomotive a good puller?". Thus far, she's identified factors which she thinks might affect the pulling power of an engine as part of her hypothesis (and whether she thinks they will have a positive or negative effect). She picked-out weight, number of motors, "grippyness" of the wheels, power applied, and diesel vs. steam. Pretty good choices, I thought.
In order to perform tests, we looked at either hooking up a bunch of cars OR building a test rig. Ultimately, the test rig won out since we will be able to try more engines in less time, and get more accurate measurements, this way. We're effectively measuring starting tractive effort.
The test rig is basically a board with powered track. Connected to the end of the board is a digital scale, capable of measuring weight (force) applied in 0.02 lb increments. The scale is, in turn, connected to whichever locomotive we are testing. We built the rig last weekend and tested it with a K-line switcher to ensure it will work. We're painting up the rig now (it's going to be pink), and hope to do our test runs this weekend. It's a Kindergarten project... so we're keeping the experiment simple, limiting the factors we'll test, testing/reporting one factor at a time, and only looking at "better or worse". We'll leave trend lines and analysis of variance to future studies 
Here's some pics of building and testing the rig:
-Dustin
