| The delta could also be within the error bounds of the measurement. Again someone who knows more about this than I could answer the questions. If were making suggestions for the next round, I'd like to propose something a long these lines. The white "x's" are measurements. 
Concentrate data and repeat measurements around the region of interest (ROI) since we can't undersize cranks any size we want. Also there are 3 repeat measurements. I don't know how expensive each measurement is, but ideally we usually take 5 repeats at different times during a test to isolate error. Also all data points should be taken randomly. A known homogeneous harness sample to take repeats from would also give you a error bound of the individual sample.
-------------------------------------------------
Work on your own car, and don't half ass it. |
