Published online by Cambridge University Press: 27 March 2009
Tests were to be made to estimate the ED 10 for surface splitting of potatoes; that is the dropping height for which on average 10% of potatoes sustain a surface split. To got good estimates for the ED 10 it is necessary to drop potatoes from heights close to this level. It was decided that a sequential procedure should be used to alter drop heights so that most drops would be made from heights close to the ED 10. Several procedures were investigated by calculation of large-sample properties of estimators and computer simulation of experiments. A procedure was then chosen for the drop tests. Results show that this procedure is easy to use, and also yields satisfactory estimates of ED 10s.