Effects of Signal-Independent Factors in Speech Perception

Abstract

The present study is an investigation of how lexical status, frequency, and neighborhood density affect the speech recognition in noise, through a replication and an extended analysis of the first experiment in Boothroyd and Nittrouer (1988). The j-factor model, proposed as a metric of context effects insensitive to overall performance level, is used to quantify the effects of these factors. All of these factors are measured with the j-factor model. The effect of neighborhood density is particularly interesting because it is primarily due to the first two segments.