And, mark, he noticed directly the wheel de- serted
and his bark off her course--and his only thought was to get that miserable, stripped, undecked, smoldering shell of a ship back again with her head pointing at her port of destination.
These tests require between 20  and 54 min  to administer to subjects with hearing impairment to identify global phoneme deficits that are significantly correlated with SeRTs, but these tests fail to permit the isolation of consonant-specific identification deficits.
In the first experiment of the current article, we described the use of the CaST as a test of consonant identification in individual subjects, focusing on the test-retest reliability and the relationship of CaST scores to SeRTs and audiometric thresholds.
We also quantified the SNRs of each syllable presented during HINT by measuring the intensities of the vowel segment of each syllable and then quantifying the SNR relative to masking noise at each SeRT. Figure 2 shows the results for 0 dB SeRTs.
HINT SeRTs averaged -1.79 dB (range -1.17 to -2.25).
Mean CaST thresholds for each subject correlated significantly with SeRTs measured with both the HINT (r = 0.62, t(14) = 3.70,p < 0.005) and the QuickSIN (r = 0.54, t(14) = 2.86, p < 0.02).
CaST thresholds accurately predicted SeRTs measured with both the HINT and the QuickSIN.
We found that only 32.5 percent of consonants could be accurately identified in isolated syllables at the SNRs that characterize SeRTs. This result agrees well with Boothroyd and Nittrouer , who reported that about 45 percent of phonemes (vowels included) could be identified in nonsense syllables at SeRTs of predictable sentences.
Group A consonants (/[??]/, /t[??]/, /t/, /s/, /z/, /d[??]/, and /r/) had average thresholds that were 2.2 dB below HINT SeRTs.
Group B consonants (/d/, /g/, /l/, /m/, /n/, /f/, and /k/) had average thresholds that were 7.4 dB above HINT SeRTs.