A study of consonant confusions in a group of cochlear impaired individuals demonstrating rollover.
Item
-
Title
-
A study of consonant confusions in a group of cochlear impaired individuals demonstrating rollover.
-
Identifier
-
AAI9325088
-
identifier
-
9325088
-
Creator
-
Dunn, Jan.
-
Contributor
-
Adviser: Harry Levitt
-
Date
-
1993
-
Language
-
English
-
Publisher
-
City University of New York.
-
Subject
-
Health Sciences, Audiology
-
Abstract
-
The present study focused on the types of consonant confusions made by cochlear-impaired listeners demonstrating rollover at intensities above (I+) and below (I{dollar}-),{dollar} the point of maximum recognition. The hypothesis to be tested stated that there would be a difference in the pattern of confusions at these two intensity levels even though the actual recognition scores were identical.;Ten cochlear-impaired subjects listened to final consonant subtests of the CUNY Nonsense Syllable Test in a cafeteria noise background at a S/N of +10. The main finding of the combined data was that voiceless consonants showed a more rapid change in performance with increasing intensity than did the voiced consonants.;Voiced consonant data showed a significant target consonant effect that approximated an ordering by manner of articulation and a vowel context effect showing greater error rates at I+ when consonants were paired with /u/. In addition, it was found that only in the case of a vowel context of /u/ did between-category errors exceed within-manner errors.;Voiceless consonant data showed a tendency for subjects to make few errors for /s/ and /sh/ while frequently missing /th/ and /f/. The effect of a context of /u/ was again evident. When grouped by relative acoustic strength, it was found that weaker consonants show greater error rates but, more importantly, there was a significant interaction of stimulus level and vowel context.;A log linear analysis of the data showed the model comprised of the response-target interaction plus the level effect plus the vowel context effect was adequate to predict the obtained confusion matrices. The response target interaction was the dominant term and, for voiced consonants, this interaction was largely explained by performance with nasal consonants.;The results were consistent with a model in which the I{dollar}-{dollar} performance could be attributed to a hearing threshold effect while the I+ performance could be attributed to an upward spread of masking effect. Alternatively, non-linear effects noted in the impaired cochlea could have produced the obtained results.
-
Type
-
dissertation
-
Source
-
PQT Legacy CUNY.xlsx
-
degree
-
Ph.D.
