Consistency and Transparency in Assessment: The Case of Grading "Participation"

In my upper-level, seminar-type courses, I generally provide, in the syllabus, something along the lines of "X% of your final grade will be based on participation" (with the rest based on papers, etc.). The X varies, but I suppose it's generally been about 30%.  (In larger and first-year courses, I do not do this, and instead say something about the possibilities of "grade bumps".)

I have to admit/confess, though, that -- even after 25 (!) years -- I have less than complete confidence in my ability to consistently, predictably, rationally "assess" "participation" and I (still) worry about the possibility that the exercise sometimes (often?) ends up as just a gussied-up bit of intuition and impression. I wonder, have others identified good ways to avert, or at least minimize, this possibility? One idea I heard from a lawprawf was to ask each student to "assess" his or her own "participation" -- to assign him or herself a letter grade for it -- and to use that self-assessment . . . in some way. Have any readers done that? Comments are open!