possibly
suggesting that subject-reported pain was a more reliable
end point for this test compared to tester-interpreted
resistance. Furthermore, mean differences between raters
were close to zero in all cases (ranging from -3 to 1)
indicating excellent agreement,'