Project Title

A comparison of word recognition performance for young adults with normal hearing when listening to two speakers

Authors' Affiliations

Madison P. Thode & Richard H. Wilson, Department of Audiology & Speech Pathology, College of Clinical and Rehabilitative Health Sciences, East Tennessee State University, Johnson City, Tennessee.

Location

WhiteTop Mountain Room 225

Start Date

4-5-2018 8:00 AM

End Date

4-5-2018 12:00 PM

Poster Number

111

Name of Project's Faculty Sponsor

Richard H. Wilson

Faculty Sponsor's Department

Department of Audiology & Speech Pathology

Type

Poster: Competitive

Classification of First Author

Graduate Student-Doctoral

Project's Category

Biomedical and Health Sciences

Abstract Text

Introduction: The purpose of the study was to analyze the differences in word recognition abilities between the Auditec (male) and VA (female) speakers for the Northwestern University Auditory Test No. 6 (NU-6) word lists for young adults with normal hearing for pure tones. Based on the literature, the hypothesis was recognition performance would be better on the Auditec version than on the VA version. The results were considered with respect to the mean data, individual subject data, and individual word data. Methods: Twelve young listeners with normal-hearing for pure tones (≤20 dB HL) participated (M = 24 years). The pure-tone thresholds were obtained using an automated test protocol. The pure-tone average (PTA) at 500, 1000, and 2000 Hz in the first session was used to reference the presentation levels of the words [-2 to 28 dB sensation level (SL) in 6-dB steps]. Each participant was presented 1200 recorded words (100 words by six presentation levels by two speakers), over three, 1-hour sessions. The NU-6 words, which were compiled with a waveform editor and in-house routines, were presented in a randomized order and at randomized presentation levels. The participants completed a questionnaire to determine the perceived difficulty level for each speaker. Results: The overall recognition performances on the Auditec and VA versions of NU-6 were 71.4% and 64.1% respectively. All subjects performed better on the Auditec version. Recognition performances on the two versions were nearly identical at higher presentation levels. The difference between the two 50% points on the overall mean psychometric functions was 3.3 dB (Auditec, 3.8-dB SL; VA, 7.0-dB SL), with equivalent slopes (4.9 and 4.8%/dB respectively). Using the Spearman- Kärber equation, the difference between the two 50% points on each of the 200 words was 2.6 dB (Auditec, M = 5.3-dB SL, SD = 3.7 dB; VA, M = 7.9-dB SL, SD = 3.7 dB), which was a significant difference. A bivariate lot of individual word performances demonstrated that Auditec had had 140 words with better performances, VA had 47 words with better performances, and 13 words were equal. The questionnaire revealed that 11 of the 12 participants indicated that the Auditec version of the NU-6 was easier to understand. Conclusions: Overall, recognition performance was better on the Auditec version but at high presentation levels, recognition performance was equivalent on both versions of NU-6. At the lower presentation levels, recognition performance was poorer with the VA speaker. It should be noted that calibration differences could account for the differences.

This document is currently not available here.

Share

COinS
 
Apr 5th, 8:00 AM Apr 5th, 12:00 PM

A comparison of word recognition performance for young adults with normal hearing when listening to two speakers

WhiteTop Mountain Room 225

Introduction: The purpose of the study was to analyze the differences in word recognition abilities between the Auditec (male) and VA (female) speakers for the Northwestern University Auditory Test No. 6 (NU-6) word lists for young adults with normal hearing for pure tones. Based on the literature, the hypothesis was recognition performance would be better on the Auditec version than on the VA version. The results were considered with respect to the mean data, individual subject data, and individual word data. Methods: Twelve young listeners with normal-hearing for pure tones (≤20 dB HL) participated (M = 24 years). The pure-tone thresholds were obtained using an automated test protocol. The pure-tone average (PTA) at 500, 1000, and 2000 Hz in the first session was used to reference the presentation levels of the words [-2 to 28 dB sensation level (SL) in 6-dB steps]. Each participant was presented 1200 recorded words (100 words by six presentation levels by two speakers), over three, 1-hour sessions. The NU-6 words, which were compiled with a waveform editor and in-house routines, were presented in a randomized order and at randomized presentation levels. The participants completed a questionnaire to determine the perceived difficulty level for each speaker. Results: The overall recognition performances on the Auditec and VA versions of NU-6 were 71.4% and 64.1% respectively. All subjects performed better on the Auditec version. Recognition performances on the two versions were nearly identical at higher presentation levels. The difference between the two 50% points on the overall mean psychometric functions was 3.3 dB (Auditec, 3.8-dB SL; VA, 7.0-dB SL), with equivalent slopes (4.9 and 4.8%/dB respectively). Using the Spearman- Kärber equation, the difference between the two 50% points on each of the 200 words was 2.6 dB (Auditec, M = 5.3-dB SL, SD = 3.7 dB; VA, M = 7.9-dB SL, SD = 3.7 dB), which was a significant difference. A bivariate lot of individual word performances demonstrated that Auditec had had 140 words with better performances, VA had 47 words with better performances, and 13 words were equal. The questionnaire revealed that 11 of the 12 participants indicated that the Auditec version of the NU-6 was easier to understand. Conclusions: Overall, recognition performance was better on the Auditec version but at high presentation levels, recognition performance was equivalent on both versions of NU-6. At the lower presentation levels, recognition performance was poorer with the VA speaker. It should be noted that calibration differences could account for the differences.