Testing methods, obviously, can vary all over the map - largely because the objectives of such tests can vary. What I'd suggest, Nicole, is to make a list of what you cant to accomplish in the tests, and then work out both a consistent provedure for exposure variations, and a way to mark the film so you'll know which roll is which later.
Your objectives, for example, might include establishing your personalized EI for specific film/developer combinations, determine push potential for specific film/developer combos (and the effects on the resulting images), etc. Defining what you want to accomplish in advance is very helpful - especially if you plan to make more precise analysis of the results with a densitometer.
Personally, I've had better luck doing exposure/development tests with static subjects (e.g. a still life setup) so that the variations in the resulting negatives can be compared more directly. If the subject changes with each frame or with different films, the differences may be difficult to recognize. Plus, it's more difficult to remember the pre-planned exposure variation sequence when you're also concerned about the subject, composition, focus, etc. I try to include items having different textures, colors, etc. in the setup of the still life, so you can get a sense of how the film handles each. I also try to include a printed grayscale step wedge in the image, so variations in contrast can be easily seen with some precision. I use a Kodak Q-14 "Color Separation Guide and Gray Scale" for this purpose. It measures about 2" x 14".
My preference is to make exposure variations in what might be called a straight line for easy comparison on the negative strips. So, if the metered exposure is M, the sequence would look like M-2, M-1, M, M+1, M+2. Again, it's much easier to do that with a static subject, where you can whip through the exposures.
But, I've also found just going out and using a new film with random shots to be useful, too. For the first roll, I use the mfgr's rated ISO, and the recommended time/temp for the developer. After developing that roll, I'll vary the assumptions based on the first-roll results. With multiple camera bodies, you can compress the cycle, but remembering which camera is which can get confusing. Removeable labels can help to keep things straight. It's also helpful to pre-make labels for the exposed rolls.
Oh, and with color negatives (assuming you'll be using a pro lab for processing), be sure to tell the lab to print using their calibration exposure and filtration, and not to adjust anything for each negative. Otherwise, they'll try to "fix" the exposure variations between individual shots.