You need to use the same film, the same exposure and two cameras side by side, with the shots taken at or near the same time.
The film needs to be slide film, and you need to have it developed by the same lab, preferably in the same run.
Try to do the test under a number of different lighting conditions. And I would aim for something like f/8 for a significant percentage of the tests, although it might be interesting to see if the colour differences diverge nearer maximum aperture.
That approach should help isolate the variables to just the difference in the lenses.