1
Taitelbaum-Swead, R., Fostick, L. (accepted, February, 2016). The effect of age and type of noise on
speech perception under conditions of changing context and noise levels. Folia Phoniatrica et
Logopaedica.
***This is a self-archiving copy and does not fully replicate the published version***
The effect of age and type of noise on speech perception under conditions of changing
context and noise levels
Riki Taitelbaum-Swead, PhD, and Leah Fostick*, PhD
Department of Communication Disorders, Ariel University, Ariel, Israel
Running Title: Age, noise, and context
*Corresponding Author:
Leah Fostick
Department of Communication Disorders
Ariel University
Ariel 40700
Israel
Tel. +972-54-4925354
Fax. +972-3-9758908
Email: [email protected]
2
3
The effect of age and type of noise on speech perception under conditions of changing
context and noise levels
Objective. Everyday life includes fluctuating noise levels, resulting in continuously
changing speech intelligibility. The study aims were (1) to quantify the amount of
decrease in age-related speech perception, as a result of increasing noise level; and (2)
to test the effect of age on context usage in word level (smaller amount of contextual
cues).
Patients and Methods. 24 young adults (age 20-30) and 20 older adults (age 60-75)
were tested. Meaningful and nonsense one-syllable consonant-vowel-consonant words
were presented with the background noise types of speech noise (SpN), babble noise
(BN), and white noise (WN), with SNR=0 and -5dB.
Results. Older adults had lower accuracy in SNR=0 with WN the most difficult
condition for all participants. Measuring the change in speech perception when SNR
decreased showed reduction of 18.6% to 61.5% in intelligibility, with age effect only
for BN. Both young and older adults used less phonemic context with WN, as
compared to other conditions.
Conclusion. Older adults are more affected by increasing noise level of fluctuating
informational noise than steady-state. They also use less contextual cues when
perceiving monosyllabic words. Further studies should take into consideration that
presenting the stimulus differently (change in noise level, less contextual cues) other
perceptual and cognitive processes are involved.
Keywords: Aging, speech perception, noise, context
4
The effect of age and type of noise on speech perception under conditions of changing
context and noise levels
INTRODUCTION
The perception and comprehension of speech is one of the basic functions of human
life. The ability to perceive speech correctly changes as a result of age and listening
conditions. Older adults frequently report difficulties in speech perception, especially
when speech is rapid or when it is accompanied by noise [1-3]. Three hypotheses
have been proposed to explain the mechanisms underlying age-related deterioration in
speech perception with background noise: peripheral hearing, central processing, and
cognitive ability [1, 4]. Indeed, various studies have showed that each of these
abilities is related to speech perception under background noise conditions [2, 3, 5-7],
and that, jointly, they can explain at least part of the variance in aging adults' speech
perception in noisy situations [8-10].
Older adults' speech perception is also affected by the type and level of noise.
The noise accompanying speech differs by the number of speakers and their gender,
and can cause energetic or informational masking. The energetic masking interferes at
the peripheral level by activating the basilar membrane at locations similar to those
stimulated by important energy in the speech wave, while the informational masking
interferes with speech processing at the central and cognitive levels. Several studies
have found that both energetic and informational masking cause greater difficulty in
speech perception for older adults than they do for younger ones [11, 12], while others
uncovered conditions under which similar interference occurs for both age groups [12,
13]. While most studies use either informational or energetic masking, few studies
5
compare the effect of both types of masking on the perception of speech in older
adults [12, 14].
Another factor that affects speech perception is the context in which it appears.
In the process of speech perception, the listener relies on bottom-up and top-down
information. When adverse listening conditions such as background noise degrade the
acoustic information, or in case of hearing loss, listeners are forced to use different
kinds of linguistic redundancy (lexical, syntactic, and semantic) and contextual cues
in order to comprehend the message [15]. Researchers have suggested that older
listeners may rely on context as a means of partially compensating for a decrease in
speech perception performance [16, 17]. Although some studies have reported that
older and younger adults use similar top-down mechanisms in a phonemic
identification task within a sentence [17, 18], others have suggested that older adults
benefit from context more than their younger peers [e.g., 15, 19, 20]. It seems that
older adults' ability to make use of contextual cues provides them with a very useful
compensatory strategy to maintain adequate language comprehension under
conditions of limited sensory information [21-23].
The literature to date describes two main methods to measure speech perception
in noise: (1) speech stimuli are accompanied with background noise with a fixed
signal-to-noise ratio (SNR) [e.g., 24]. The individual's speech perception ability is
then defined by the accuracy rate of identifying speech at this SNR level; (2)
adaptively changing background noise, to identify the level of noise required for the
participants to perceive correctly the speech stimuli [e.g., 12, 25]. The outcome of
both methods reveals the individual's ability to perceive speech in a specific, fixed,
noise level. However, in day to day life, noise level varies continuously during
speech, creating endlessly changing conditions while perceiving speech. There is no
6
data showing how much one's ability to perceive speech-in-noise decreases when
noise level increases. Furthermore, there is no data as to how increasing noise levels
affect older adults as compared with younger ones, and how the effect differs between
different types of noise.
Hence, the aim of the current study was to quantify the effect of increasing
noise level (from SNR=0 to SNR=-5 dB) on speech perception accuracy, and to
measure whether this effect changes as a function of age group (young vs. older
adults) and type of noise (speech noise [SpN], babble noise [BN], and white
noise[WN]). In addition, we tested the contribution of lexical context (meaningful vs.
nonsense words, and perception of word-parts vs. whole word) to speech perception
in different age groups and under varying background noise conditions.
7
METHOD
Participants
The study included 44 Hebrew-speaking adults, of whom 24 were young adults
between the ages of 20 and 30 years and 20 were older adults, aged 60-75 years. All
participants were screened for age-normal hearing at 500, 1,000, 2,000, and 4,000 Hz,
according to Lebo and Redell [26]. Six additional older adults were recruited but
failed the screening, and were therefore excluded from the study.
Task and stimuli
Speech perception for meaningful words was tested using the Hebrew version
of the AB words test [27]. This test is composed of lists of ten meaningful, onesyllable Consonant-Vowel-Consonant words, which are phonetically balanced (i.e., in
each list, every consonant appears once and every vowel appears twice). Speech
perception for nonsense words was tested using similar phonetically balanced lists of
ten one-syllable Consonant-Vowel-Consonant nonsense words. The words were
separated by silent gaps of three seconds' duration, and the participants were required
to repeat each word immediately after hearing it during these gaps.
The words were presented in 12 different conditions: two context conditions
(meaningful or nonsense words), three background noise conditions (BN, SpN, or
WN), and two noise/SNR level conditions (SNR of 0 or -5 dB). Two lists (20 words)
were presented in each condition. Meaningful and nonsense words were presented
separately in different experimental blocks, and the participants were informed that
they would hear short words, either meaningful or nonsense. In order to avoid effects
related to the order of the lists and the words within lists, the assignment of each list
8
to any of the conditions, the lists’ order of presentation, and the order of the words
within each list were randomly intermixed. For each condition, the number of
correctly recognized words (out of 20) and phonemes (out of 60, three for each word)
was recorded.
The target words were presented by a female native Hebrew speaker. SpN and
WN were generated using the Diagnostic Audiometer DA64, and demonstrated
identical spectrotemporal features in every trial. SpN was composed of frequencies
within the range of 0.5-2 kHz (band-pass noise) with constant amplitude. WN was
composed of wide-band noise, which was evenly distributed over the entire range of
frequencies (.25-8 kHz). BN was composed of four native Hebrew speakers, two
males and two females, whose speech was normalized to have the same RMS
amplitudes and no breaks within words. Each of the speakers read a different,
meaningful passage. The silent gaps between the words and sentences were deleted,
creating continuous streams of words from all speakers. All noises were added to the
words at SNR levels of 0 and -5 dB. The noise started three seconds before the first
word in the list and continued non-stop through the list, ending with the last word.
The words’ level was normalized using the overall RMS. Words were presented at a
rate of 120 per minute and were delivered binaurally at a level of 35 dB SL by the
audiometer, via headphones.
Apparatus
The words were recorded in a recording studio using a SONTRONICS TCS-6
microphone and Samplitude classic 8.1 recording software. They were edited using
the SoundForge program, which digitized (16-bit) at a sampling rate of 44 kHz. The
9
SoundForge program was also used to compose the BN and to add the different
background noises to the words.
The words were presented using the Winamp Media Player 5.7 software, via STech supra aural headphones. Participants' responses were recorded using a SONY
ICD-PX312 recording device, which was placed in close proximity to the participants.
The Interacoustic AD229B audiometer was used to screen and measure hearing level.
Procedure
The study was approved by the ethics Institution Review Board (IRB) and was
conducted in accordance with Good Clinical Practice (GCP) guidelines [28, 29]. All
participants received full explanation about the study and signed an informed consent.
All potential participants were screened for hearing level prior to participation in the
study, and for all participants that completed screening, this test was used to measure
their exact hearing level so as to determine the words’ presentation level during the
experiment (35 dBSL).
The participants' responses were recorded, and two independent raters
transcribed the words they said. Where the raters disagreed (this occurred in less than
1% of the words), a transcription of a third rater was used.
10
RESULTS
A. Effect of age and type of noise on speech perception accuracy
Figure 1 presents the main data of speech perception accuracy for meaningful
words with SNR=0 for different noise conditions and age groups. Repeated measures
Analysis of Variance (ANOVA) for raw accuracy data revealed significant main
effects for age (F(1,42)=35.067, p<.001) and type of noise (F(2,84)=19.165, p<.001). As
expected, speech perception accuracy was higher for young than for older adults in all
noise conditions. A significant age by noise type interaction (F(2,84)=3.33, p<.05) was
observed. Although all three post-hoc tests were significant (for WN: t(42)=2.295,
p<.05, for SpN: t(42)=5.173, p<.001, for BN: t(42)=5.271, p<.001) and in the same
direction (i.e., younger subjects with better scores than older subjects), the age related
differences are larger for babble and speech noise as compared to white noise, as
shown in Figure 1; this means the overall interaction effect is relatively weak. .
****PUT FIGURE 1 AROUND HERE***
B. Effect of age and type of noise on changes in speech perception accuracy due
to worsening noise and context conditions
1. Noise level: Effect of decrease in SNR on speech perception accuracy
In order to measure the change in speech perception accuracy as a result of
increasing noise level from SNR=0 to SNR=-5 dB, the percentage of change in
speech perception accuracy from SNR=0 to SNR=-5 dB was calculated as {[(SNR(0)SNR(-5)]/SNR(0)}. Repeated Measures ANOVA was used, with type of noise (SpN,
BN, WN) and context (nonsense words, meaningful words) as within-subject
variables, and age group (young, older adults) as a between-subject variable.
11
The percentage of change in speech perception for young and older
participants in all the study conditions is presented in Table 1. Contrary to the results
on raw accuracy scores, young and older participants did not differ in the overall
percentage of change in speech perception accuracy, and both groups had a mean
reduction in speech perception accuracy of 33 – 39% (s.d. = .031-.034, F(1,41)=1.783,
p>.05, see table 1). However, significant effects were found both for type of noise
(F(2,84)=9.634, p<.001) and for context (F(1,41)=13.783, p<.01). As can be expected, a
larger decrease was found in nonsense words’ intelligibility as a result of increasing
noise level as compared to meaningful words. Interestingly, the increase in noise level
from SNR=0 to SNR=-5 dB resulted in a greater decrease of accuracy in BN than it
did in either SpN (p<.001) or WN (p<.01). No difference was found between SpN and
WN. Noise by age group interaction (F(2,84)=4.891, p<.05) revealed a greater decrease
in accuracy for older than for young participants in BN, but no difference between age
groups in SpN and WN (Table 1). No context by age group or noise by context
interactions were found (F(1,41)=.292, p>.05 and F(2,82)=2.655, p>.05, respectively).
****PUT TABLE 1 AROUND HERE***
2. Context I: lexical context’s contribution to speech perception accuracy
Differences in the perception of nonsense vs. meaningful words were calculated
using the formula of Boothroyd and Nittrouer [16] and Nittrouer and Boothroyd [26]
for calculating the k-factor for speech perception accuracy with SNR=0 for nonsense
words and meaningful words, as follows:
k = log (1-pc) / log (1-pi)
12
Where pc is the probability of recognition in context, and pi is the probability of
recognition with no context.
A Repeated Measures ANOVA was carried out on k-factor scores, with
different noise types (SpN, BN, WN) as a within-subject variable, age group (young,
older adults) as a between-subject variable. No main effect was found for type of
noise (F(2,68)=.990, p>.05) and age (F(1,34)=.817, p>.05), nor any interaction,
suggesting that the relation between perceiving nonsense and meaningful words was
similar in both age groups and types of noise.
3. Context II: Perception of parts vs. whole
An additional aspect of context usage that was explored was the participants'
ability to perceive fragments of the words vs. the whole words. The formula of
Boothroyd and Nittrouer [16] and Nittrouer and Boothroyd [26] for calculating the jfactor was applied for speech perception accuracy with SNR=0 for phoneme and
words accuracy, as follows:
j = log (pw) / low (pp),
Where pw is the probability of recognition of the whole word, and pp is the probability
of recognition each of its parts [16, 26].
Repeated Measures ANOVA was carried out on j-factor scores, with different
noise type (SpN, BN, WN) and lexical context (nonsense words, meaningful words)
as within-subject variables, age group (young, older adults) as a between-subject
variable. Significant main effects were found for type of noise (F(2,68)=4.576, p<.05)
and lexical context (F(1,34)=19.124, p<.001), but not for age group (F(1,34)=.001,
p>.05). Larger j-factors were observed, i.e., more phonemes were needed, for speech
perception accompanied by WN (mean= 3.189, s.d.=.097) than for SpN (mean=
13
2.904, s.d.=.075, p<.05) and BN (mean= 2.798, s.d.=.106, p<.01). No difference was
found between SpN and BN. As can be expected, larger j-factors, with more
phonemes for identifying words, were needed for nonsense words (mean=3.171,
s.d.=.069) in comparison to meaningful words (mean=2.756, s.d.= .073, p<.001). Age
group by context interaction (F(1,34)=6.314, p<.05) revealed no differences between
age groups in meaningful words (t(42)=-.928, p>.05), and larger j-factors for young
participants in nonsense words (mean=3.293, s.d.=.527), as compared to than older
adults (mean=3.050, s.d.=.235, t(42)=1.996, p<.05). However, while this difference is
statistically significant, it does not carry any clinical meaning. Since the words were
one-syllable CVC words, with only three phonemes in each word, a j-factor larger
than three is meaningless. Therefore, as both groups had a j-factor of three, we
considered both groups equivalent, i.e., no differences were observed between age
groups in j-factor.
14
DISCUSSION
Understanding speech in daily life is a complicated task, especially so since we
perform it under ever-changing listening conditions. These changing conditions occur
on the backdrop of relatively ‘constant’ characteristics of the listener (such as age)
and the environment (such as the type of noise). The current study aimed to test the
effect of age and type of noise on speech perception in different noise level and
context conditions. As was expected, speech perception accuracy was lower for older
participants than it was for younger ones in all conditions. In addition, speech
perception under WN generally proved to be the most difficult, harder than both under
BN and SpN.
The performance of older adults under different noise conditions was previously
studied [e.g., 30, 31], and our finding of WN being the most difficult condition
comply with them [e.g., 1, 32]. The increase in noise level caused a decrease in
speech perception, depending on age group, noise, and context conditions (Table 1).
Testing the effect of the change in noise level, rather than the presence of noise per se,
revealed interesting findings regarding BN, both in general and when comparing age
groups. In general, the increase in noise level (from SNR=0 to SNR=-5 dB) resulted
in a greater decrease in speech perception in BN, as compared with SpN and WN
which had similar effect for the increase in noise level, probably since both are
energetic maskers. However, BN contains both energetic and informational masking
which might cause larger interference with speech intelligibility, as compared with
only energetic maskers [12, 14].
In terms of age differences, the change in speech perception accuracy under BN
was larger among older vs. young adults. Indeed, studies suggest that difficulties for
15
older participants emerge only when informational masking is present [32]. The fact
that SpN and BN had the same energetic levels, but that the increase in noise level for
BN affected older participants more than it did young ones, supports Rajan &
Cainer’s [12] suggestion that the greater effect of BN on older participants evolves
from a decrease in cognitive ability for older participants, who encounter an increase
in cognitive load as a result of the informational masking.
The literature suggests that older adults use more context cues than younger
ones [15-17], but most of these studies used sentences as the target stimulus [e.g., 26,
33, 34]. In the current study, we were interested to see whether the pattern that older
adults use more contextual cues than younger ones applies also when speech
perception is measured at the word level. We were also interested to find out how
different types of noise affect speech perception at the word level, which was not
studied to date. We therefore used one-syllable CVC words, and measured context as
the amount of information in the word used by the listener in order to understand it
correctly.
The type of masker, namely the amount of masking energy, affected the number
of phonemes required for perceiving words, resulted in larger j-factor for WN, while
the narrow-band maskers (BN and SpN) had lower and similar j-factors. This is the
first study to measure k- and j- factors in different background noise conditions, and it
suggests that the frequency range of the masker is the factor affecting j-factor, and not
the energetic or informational characteristics. For nonsense words, participants
needed to hear the whole word (j=~3) in order to perceive it correctly. Similar
findings were reported by Boothroyd and Nittrouer [16] and by Nittrouer and
Boothroyd [26]. Most importantly, no effect for age on both k- and j-factors suggest
that older adults’ larger use of contextual cues, is specific to speech stimuli that
16
contains several types of contextual cues (i.e., lexical, syntactic, and semantic), as in
the case of sentences, and does not apply to stimuli in the word level.
To sum up, in the current study we quantified young and older participants'
changes in speech perception, as a result of increasing noise level, for words with
different background noises. This quantification revealed a reduction of 18.6% to
61.5% in speech perception, depending on age group, noise, and context conditions.
The findings show that older participants are disadvantaged with regard to speech
perception in noisy conditions. Increasing noise level had different effects on different
types of noise, and with different age groups. The effect of some noises on speech
intelligibility will increase linearly with noise level, while for others it will increase
exponentially. Theoretically, the fact that increasing noise level resulted in larger
decrease in speech intelligibility in BN, than in SpN or WN, although WN was the
most difficult condition when fixed SNR=0, suggests that speech accompanied by
fluctuating background noise is more prone to be affected than speech in steady-state
noise [12, 14]. This should be further studies with additional types of noise and be
considered clinically, when testing older adults' speech intelligibility in background
noise. In addition, we found that older adults' larger use of contextual cues applies
only to the sentence level when all kinds of cues are available (such as lexical,
syntactic and semantic cues). However, when the availability of these cues is reduced,
and only lexical cues are available, along with some semantic information, there is no
age effect. This could have resulted from a "floor effect" in contextual cues when
using word level. It also might suggest that older adults' larger use of contextual cues
is only for syntactic and semantic cues, and not for lexical cues. The latter imply that
higher cognitive processes are involved when aging adults use contextual cues. Aging
17
effect of different types of contextual cues should be further studied, and taken into
consideration in clinical testing of speech perception.
18
ACKNOWLEDGMENTS
The authors declare no conflict of interests.
The study was not supported by any sponsor or funding agency.
19
REFERENCES
1. Fostick L, Ben-Artzi E, Babkoff H. Aging and speech perception among the
elderly: Beyond hearing threshold and cognitive ability. J Basic Clin Physiol
Pharmacol 2013;24:175-183.
2. Schneider BA, Daneman M, Pichora-Fuller MK. Listening in aging adults: From
discourse comprehension to psychoacoustics. Can J Exp Psychol 2002;56:139152.
3. Wingfield A, Lindfield KC, Goodglass H. Effects of age and hearing sensitivity on
the use of prosodic information in spoken word recognition. J Speech Lang Hear
Res 2000;43:915-925.
4. Committee on Hearing, Bioacoustics and Biomechanics (CHABA). Speech.
Understanding and Aging. J Acoust Soc Am 1988;83:859-820.
5. Frtusova JB, Winneke AH, Phillips NA. ERP evidence that auditory-visual speech
facilitates working memory in younger and older adults. Psychol Aging
2013;28:481-494.
6. Grossman M, Cooke A, Devita C, et al. Age-related changes in working memory
during sentence comprehension: an fMRI study. Neuroimage 2003;15:302-317.
7. Humes LE, Dubno JR, Gordon-Salant S, et al. Central presbycusis: A review and
evaluation of the evidence. J Am Acad Audiol 2012;23:635-666.
8. Anderson MC, Arehart KH, Kates JM. The effects of noise vocoding on speech
quality perception. Hear Res 2014;309:75-83.
9. Harris KC, Dubno JR, Keren NI, et al. Speech recognition in younger and older
adults: a dependency on low-level auditory cortex. J Neurosci 2009;29:60786087.
20
10. Schvartz KC, Chatterjee M, Gordon-Salant S. Recognition of spectrally degraded
phonemes by younger, middle-aged, and older normal-hearing listeners. J Acoust
Soc Am 2008;124:3972-3988.
11. Helfer KS, Staub A. Competing speech perception in older and younger adults:
behavioral and eye-movement evidence. Ear Hear 2014;35:161-170.
12. Rajan R, Cainer KE. Ageing without hearing loss or cognitive impairment causes
a decrease in speech intelligibility only in informational maskers. Neuroscience
2008;154:784-795.
13. Agus TR, Akeroyd MA, Gatehouse S, et al. Informational masking in young and
elderly listeners for speech masked by simultaneous speech and noise. J Acoust
Soc Am 2009;126:1926-1940.
14. Ezzatian P, Li L, Pichora-Fuller K, et al. The effect of priming on release from
informational masking is equivalent for younger and older adults. Ear Hear
2011;32:84-96.
15. Rogers CS, Jacoby LL, Sommers MS. Frequent false hearing by older adults: The
role of age differences in metacognition. Psychol Aging 2012;27:33–45.
16. Boothroyd A, Nittrouer S. Mathematical treatment of context effects in phoneme
and word recognition. J Acoust Soc Am 1988;84:101-114.
17. Dubno JR, Ahlstrom JB, Horwitz AR. Use of context by young and aged adults
with normal hearing. J Acoust Soc Am 2000;107:538-546.
18. Saija JD, Akyürek EG, Andringa TC, et al. Perceptual restoration of degraded
speech is preserved with advancing age. J Assoc Res Otolaryngol 2014;15:139148.
21
19. Fogerty D, Kewley-Port D, Humes LE. The relative importance of consonant and
vowel segments to the recognition of words and sentences: effects of age and
hearing loss. J Acoust Soc Am. 2012;132(3):1667-78.
20. Rogers CS, Jacoby LL, Sommers MS. Frequent false hearing by older adults: the
role of age differences in metacognition. Psychol Aging. 2012;27(1):33-45.
21. Baum SR. Age differences in the influence of metrical structure on phonetic
identification. Speech Commun 2003;39:231-242.
22. Wingfield A. Cognitive factors in auditory performance: Context, speed of
processing, and constraints of memory. J Am Acad Audiol 1996;7:175-182.
23. Wingfield A, Wayland SC, Stine EA. Adult age differences in the use of prosody
for syntactic parsing and recall of spoken sentences. J Gerontol 1992;47:350-356.
24. Souza PE, Boike KT, Witherell K, Tremblay K. Prediction of speech recognition
from audibility in older listeners with hearing loss: effects of age, amplification,
and background noise. J Am Acad Audiol 2007;18(1):54-65.
25. Frisina DR, Frisina RD. Speech recognition in noise and presbycusis: Relations to
possible neural mechanisms. Hear Res 1997;106:95-104.
26. Lebo CP, Reddell RC. The presbycusis component in occupational hearing loss.
Laryngoscope 1972;82:1399-1409.
26. Nittrouer S, Boothroyd A. Context effects in phoneme and word recognition by
young children and older adults. J Acoust Soc Am 1990;87:2705-2715.
27. Boothroyd A. Statistical theory of the speech discrimination score. J Acoust Soc
Am 1968;43:362-367.
28. Kolman J, Meng P, Graeme S. Good clinical practice. Standard Operating
Procedure for Clinical. 1998.
22
29. Grimes DA, Hubacher D, Nanda K, Schulz KF, Moher D, Altman DG. The Good
Clinical Practice guideline: a bronze standard for clinical research. The Lancet.
2005;366(9480):172-4.
30. Brungart DS, Sheffield BM, Kubli LR. Development of a test battery for
evaluating speech perception in complex listening environments. J Acoust Soc
Am. 2014;136(2):777-90.
31. Jin SH, Liu C, Sladen DP. The effects of aging on speech perception in noise:
comparison between normal-hearing and cochlear-implant listeners. J Am Acad
Audiol. 2014;25(7):656-65.
32. Tun A, Wingfield A. One Voice Too Many: Adult Age Differences in Language J
Gerontol B Psychol Sci Soc Sci 1999; 54B(5):317-327.
33. Uslar V, Ruigendijk E, Hamann C, et al. How does linguistic complexity
influence intelligibility in a German audiometric sentence intelligibility test? Int J
Audiol 2011;50:621–631.
34. Woods DL, Doss Z, Herron TJ, et al. Age-related changes in consonant and
sentence processing. J Rehabil Res Dev 2012;49:1277-1291.