Abstract
To examine the repeatability of classification of the cleanness test for raw silk, a computer simulation experiment was carried out basedonrepeatedtestingofthe same lot. The differences in applicable grade of cleanness were mostly the same or grade 1, but differences in grade 2 appeared at the rate of 8% for 5A, 19% for 4A, 9% for 3A and 5% for 2A. Differences in grade 3 appeared at 3% for 5A and 1% for 2A and A.
The probability of seeing a difference when grading twice for cleanness was 60% for 5A, 65% for 4A, 69% for 3A, 39% for 2A and 36% for A. There is a marked difference between the experimental value for the probability of a difference appearing when grading twice and the calculated value based on the probability of grading errors. (E mail : kouzu_ub@nifty.com)
