Reprinted from Annals of Otology, Rhinology

INTRODUCTION When summarizing the performance achieved by users of a cochlear prosthesis, results are typically reported as means and standard deviations of speech recognition scores1 (also Brimacombe and Beiter, unpublished observations, 1992). These statistics constitute a valid description of the data only when scores are distributed normally, ie, as a bell-shaped On the basis of theoretical considerations, however, a bimodal distribution might be expected. Patients with poor auditory nerve survival or atrophic central auditory systems would be expected to have uniformly poor results, while patients with intact auditory nervous systems should, in general, achieve excellent results with a well-designed and -fitted cochlear prosthesis. The Clarion cochlear prosthesis provides both the Multi-Strategy speech processof' and the University of CaliforniaSan Francisco (UCSF) radial bipolar electrode: a combination that provides capabilities to fit the stimulation strategy to the auditory nervous system of each individual patient. We have examined the distributions of speech recognition scores in order to understand the relative importance of biological versus technological limits in the benefits derived from cochlear prostheses.