Stimulus Ratio Effects on Speech Discrimination by Children and Adults

This investigation examined the response strategies and discrimination accuracy of adults and children aged 5–10 as the ratio of same to different trials was varied across three conditions of a "change/no-change" discrimination task. The conditions varied as follows: (a) a ratio of one-third same to two-thirds different trials (33% same), (b) an equal ratio of same to different trials (50% same), and (c) a ratio of two-thirds same to one-third different trials (67% same). Stimuli were synthetic consonant-vowel syllables that changed along a place of articulation dimension by formant frequency transition. Results showed that all subjects changed their response strategies depending on the ratio of same-to-different trials. The most lax response pattern was observed for the 50% same condition, and the most conservative pattern was observed for the 67% same condition. Adult response patterns were most conservative across condition. Differences in discrimination accuracy as measured by P(C) were found, with the largest difference in the 5- to 6-year-old group and the smallest change in the adult group. These findings suggest that children's response strategies, like those of adults, can be manipulated by changing the ratio of same-to-different trials. Furthermore, interpretation of sensitivity measures must be referenced to task variables such as the ratio of same-to-different trials.

KEY WORDS: developmental speech perception, school-aged children, a priori proportion (of stimulus types)

Submitted on November 20, 1989
Accepted on August 24, 1990

CiteULike    Connotea    Del.icio.us    Digg    Facebook    Reddit    Technorati    Twitter    What's this?