ABSTRACT
This study examined the interaction of acoustic-phonetic information with higher-level linguistic contextual information during the real-time speech perception process in child, young adult, and older adult listeners. Five age groups were studied: (a) young children ranging in age from 5 to 7 years, (b) older children aged 8 to 10 years, (c) young adults aged 18 to 23 years, (d) older adults aged 60 to 69 years, and (e) older adults aged 70 to 83 years. All subjects were presented with time-gated monosyllabic target words presented in sentence contexts containing contrasting levels of word predictability. Findings indicated that target word predictability influenced the timing and nature of the real-time recognition process including the listeners' use of initial word sounds. Predictability-high (PH) words were recognized earlier and with greater confidence than predictability-low (PL) words. PH recognition performance was more influenced by child development and aging than PL recognition performance. Older adult listeners required more PH-gated word stimuli to produce accurate responses than younger adults. Older children showed more effective use of PH contexts than younger children.
Subject(s)
Aging , Child Language , Language , Speech Perception , Speech , Vocabulary , Acoustic Stimulation , Adult , Age Factors , Aged , Audiometry, Pure-Tone , Child , Child, Preschool , Female , Hearing , Humans , Language Development , Male , Middle Aged , Time FactorsABSTRACT
This study examined the effects of signal presentation level and word duration on time-gated isolated monosyllabic word-recognition performance. Measures of listener confidence, word identification, isolation point (IP), confidence at IP, and acceptance point were obtained from normal-hearing listeners. Subjects were presented with non-time-gated and time-gated speech stimuli at 40 dB SPL (N = 21). The resulting performance measures were compared with previously reported results obtained using an 80-dB SPL presentation level. The speech stimuli consisted of 60-msec time-gated isolated monosyllabic words developed from a prerecorded 50-item list (Auditec, NU-6). Comparisons were drawn between presentation levels, word durations, and time-gated and non-time-gated presentation conditions. Poorer accuracy and longer isolation points were observed at the lower signal presentation level. The findings further indicated that listener confidence at IP was only indirectly influenced by presentation level. Monosyllabic word duration was a significant factor in on-line recognition performance, regardless of presentation level.
Subject(s)
Speech Acoustics , Speech Discrimination Tests/standards , Speech Perception/physiology , Adult , Evaluation Studies as Topic , Female , Humans , Psychoacoustics , Speech Discrimination Tests/psychology , Time FactorsABSTRACT
This investigation was designed to study real-time isolated monosyllabic word-recognition performance and the feasibility of applying time-gated NU-6 word-recognition test materials for real-time assessment of older listeners. Methods and materials developed in a previous investigation were used to obtain time-gated performance measures from 37 older listeners (mean age = 69 years). The older listener performance measures were compared with extant data from 20 normally hearing young adult listeners (mean age = 22 years). Specifically, listener confidence and accuracy by gate as well as listener isolation point, confidence at the isolation point, and total acceptance point measures were evaluated. The results show that major events in the real-time understanding process occur at a slower pace among older listeners. The data indicate that the time-gating method has excellent potential for future research among elderly listeners.
Subject(s)
Aging/psychology , Audiometry, Speech/methods , Speech Perception , Adult , Aged , Female , Humans , Male , Middle Aged , Speech Discrimination Tests , Time FactorsABSTRACT
This study examined the effects of forward time gating and word length on monosyllabic isolated word-recognition performance. Fifty (60-ms) time-gated words were developed from a pre-recorded version (Auditec of St. Louis) of the Northwestern Auditory Test No. 6 (NU-6) List 1, Form A. A total of 358 time-gated items were presented monaurally at 80 db SPL, and time-gated word identification, isolation point, acceptance point, and listener confidence measures were obtained from 20 normally hearing listeners. A comparison of conventional nontime-gated word-recognition scores obtained using the NU-6 List 4, Form C with the time-gated measures revealed that, even upon word offset, listeners recognized time-gated words less frequently and with less confidence. The time-gated findings also demonstrated that word length, based on word duration from onset to offset, significantly influenced real-time recognition performance.
Subject(s)
Phonetics , Speech Discrimination Tests/standards , Speech Perception/physiology , Adult , Evaluation Studies as Topic , Female , Humans , Models, Biological , Reaction Time , Time FactorsABSTRACT
This study compared the effects of three conditions of word predictability on recognition performance. Word-recognition scores were obtained for 30 normally hearing listeners using three pairs of sentence lists. Each list contained target words preceded by a variety of predictability-high (PH) and predictability-low (PL) phrases and a constant predictability-neutral carrier phrase (CP) in three signal-to-babble ratio (S/B) listening conditions (-5, 0, and +5 dB). Significant differences (p less than .05) were observed between PH, PL, and CP word-recognition scores for specific S/B's and pairs of sentence lists. The results were interpreted in terms of cognitive processing tasks associated with the various semantic and contextual cues provided by each condition of word predictability.
Subject(s)
Speech Discrimination Tests , Speech Perception , Adult , Cues , Female , Humans , Male , Speech IntelligibilityABSTRACT
Word discrimination scores were obtained for 30 normally hearing listeners using Auditec and Rintelmann recordings of the NU-6 presented at 60 and 70 dB SPL in quiet and in a background of multitalker noise at signal-to-noise (S/N) ratios of +12, +6, 0, and -6 dB. This was done to obtain normative data and to determine whether these two recordings of the same word lists were equivalent at conversational levels in quiet and in a noisy background. The word discrimination scores for each recording in quiet and at each S/N ratio were not significantly different when presented at 60 and 70 dB SPL. However, the word discrimination scores were significantly poorer as the S/N ratio became more severe, and the scores obtained with the Auditec recording were significantly poorer than those obtained with the Rintelmann recording at each S/N ratio. These differences between word discrimination scores using the Auditec and Rintelmann recordings were primarily attributed to the recorded level of the test word and secondarily attributed to carrier phrase and talker differences.
Subject(s)
Speech Discrimination Tests , Speech Perception , Adult , Calibration , Female , Humans , Male , Noise , Reference Values , Speech Discrimination Tests/instrumentation , Speech Discrimination Tests/methodsABSTRACT
Monaural and binaural SRT and speech discrimination scores in a sensorineural hearing-imparied sample (N: 194, aged 3--91 yrs, mn=43 yrs) indicated that while a majority of clients have neither a binaural advantage or disadvantage by these tests, an important minority (5% for SRT, 20% for discrimination) have worse binaural than better-ear scores. Binaural as well as monaural tests of SRT and of speech discrimination should be a standard part of a hearing aid evaluation.
