As long as you have a good measurement system like a microscope, why not use it
to measure something with higher information content rather than the operator's
opinion. Scratch length, scratch depth, scratch area, scratch volume, stain
area, stain depth, pit size...you get the idea. You will be much better off in
the long run using a measure with more information content than grading or
binary classification. Gauge R&R should include important predicted effects
such as lighting changes, operator fatigue, defocus, etc if you expect to get
a reasonable estimate of measurement error.
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
