J Am Acad Audiol 25:605-623 (2014)

Loudness as a Cue for Acceptable Noise Levels DOI: 10.3766/jaaa.25.6.10 Karrie L. Recker Martin F. McKinney Brent W. Edwards

Abstract Background/Purpose: The acceptable noise level (ANL) test is the only test that is known to predict success with hearing aids with a high degree of accuracy. A person’s ANL is the maximal amount of background noise that he or she is “willing to put up with” while listening to running speech. It is defined as the speech level minus the noise level, in decibels (dB). People who are willing to put up with high levels of background noise are generally successful hearing-aid wearers, whereas people who are not willing to put up with high levels of background noise are generally unsuccessful hearing-aid wearers. If it were known what cues that listeners are using to decide how much background noise they are willing to tolerate, then it might be possible to create technology that reduces these cues and improves listeners’ chances of success with hearing aids. As a first step toward this goal, this study investigated whether listeners are using loudness as a cue to determine their ANLs. Research Design and Study Sample: Twenty-one individuals with normal hearing and 21 individuals with sensorineural hearing loss participated in this study. In each group of 21 participants, 7 had a low ANL (< 7 dB), 7 had a mid ANL (7-13 dB), and 7 had a high ANL (>13 dB). Data Collection/Analysis: Participants performed a modified version of the ANL in which the speech was fixed at four different levels (50,63,75 and 88 dBA), and participants adjusted the background noise (multitalker babble) to the maximal level at which they were willing to listen while following the speech. These results were compared with participants’ equal-loudness contours for the multitalker babble in the presence of speech. Equal-loudness contours were measured by having the participants perform a loudness-matching task in which they matched the level of the background noise (multitalker babble), played concurrently with speech, to a reference condition (also multitalker babble). During the test condition, the speech played at 50, 63, 75, or 88 dBA. All testing was performed in a sound booth with the speech and the noise presented from a loudspeaker at a 0° azimuth, 3 feet in front of the participant. Each condition was presented multiple times, and the results were averaged. Presentation order was randomized. Participants were tested unaided. Results: Participants’ ANLs were compared with their equal-loudness contours for the background noise. ANLs that ran parallel to the equal-loudness contours were considered consistent with a loudness-based listening strategy. This pattern was observed for only two participants - both hearingimpaired. Conclusions: The majority of listeners showed no consistent trend between their ANLs and their loudness-matched data, suggesting that they are using cues other than loudness to determine their ANLs. ANLs were consistent with loudness-matched data for a small subset of listeners, suggesting that they may be using loudness as a cue for determining their ANLs. Key Words: Acceptable noise level, loudness Abbreviations: ANL = acceptable noise level; BNL = background noise level; CD = compact disk; MCL = maximal comfortable level; MinASL = minimal acceptable speech level; SNR = signal-to-noise ratio; SPL = sound pressure level

Starkey Hearing Technologies, Eden Prairie, Karrie Recker, Starkey Hearing Technologies, 6600 Washington Ave. S., Eden Prairie,

I 55344; E-mail: [email protected]

m l™ r n n ? ere Presen^ d at the Audiology Association, February 17,2010, Bogota, Columbia; American Academy of Audiology, April 14 2010, San Diego CA International Hearing Aid Research Conference, August 16, 2010, Lake Tahoe, CA; University of Minnesota December 3 2010 Minneapolis, MN; and the World Congress of Audiology, May 2, 2012, Moscow, Russia. ’ ecemDerJ’ ^ U1U'

605

Journal o f the American Academy o f Audiology /Volume 25, Number 6, 2014

INTRODUCTION he acceptable noise level (ANL) test is the only test known to predict success with hearing aids with a high degree of accuracy (Nabelek et al, 2006). This test is a measure of the maximal amount of background noise that a listener is “willing to put up with” while listening to running speech (Nabelek et al, 1991). ANLs are measured by having a listener adjust running speech to a comfortable level. Next, background noise is added and the listener adjusts its level to the m axim al level that he or she is willing to put up with while listening to the speech. The ANL equals the max­ imal comfortable level (MCL) for the speech minus the maximal tolerable background noise level (BNL), in dec­ ibels (dB). For the traditional method of ANL testing, in which both the speech and the noise are presented dioticly (from a loudspeaker at a 0° azimuth relative to the listener or over circumaural headphones/insert earphones), most people have an ANL between -5 and 30 dB (Rogers et al, 2003; Mueller et al, 2006; Nabelek et al, 2006; Tampas and Harkrider, 2006; Recker et al, 2011). Other presentation methods (e.g., dichotic) have been shown to yield different ranges (e.g., Peeters et al, 2009). Generally, people who are willing to accept high lev­ els of background noise (those who have ANLs of ^ 7 dB) tend to be successful hearing-aid wearers, whereas people who are not willing to accept high levels of background noise (those who have ANLs of > 13 dB) tend to be unsuc­ cessful hearing-aid wearers (Nabelek et al, 2006). People who are willing to accept moderate levels of background noise (those who have ANLs 7-13 dB) may or may not be successful with hearing aids. For the American version of the test, ANLs are known to be stable for several months’ time (Nabelek et al, 2004). However, repeatability has been shown to be much poorer with the Danish version of the test (Olsen et al, 2012a,b). Several studies have tried to correlate listeners’ ANLs with other measures. Generally, they have had minimal success. ANLs do not appear to correlate with listeners’ age (Nabelek et al, 1991), gender (Rogers et al, 2003), hearing sensitivity (Nabelek et al, 1991), the type of background noise (Lytle, 1994; Crowley and Nabelek, 1996) - although ANLs obtained with music as a back­ ground noise are different from the ANLs for other types of background noise (Nabelek et al, 1991; Gordon-Hickey and Moore, 2007) - speaker gender, or the interest level of the speech material (Plyler et al, 2011). Additionally, investigations into specific attributes of communication using the Abbreviated Profile of Hearing Aid Benefit questionnaire found no significant correlation between ANLs and any of the subscales th at it measures: ease of communication, reverberation, background noise, and aversiveness to sounds (Freyaldenhoven et al, 2008a).

T

606

However, two of these subscales (ease of communication and background noise), when used in conjunction with ANLs, improved the accuracy of predicting who would be a successful hearing-aid w earer by 6% (range 8591%). Studies th at have tested ANLs with the speech pre­ sented at multiple fixed levels, rather than at the MCL, have generally found th at as the level of the speech increases, ANLs also increase (Franklin et al, 2006; Tampas and Harkrider, 2006; Freyaldenhoven et al, 2007; Recker and Edwards, 2013). However, this is not true for all individuals. In Recker and Edwards (2013), 20% of the norm al-hearing and 20% of the hearingimpaired participants chose to listen at a constant ANL across a 38 dB range of speech levels (50-88 dBA). Freyaldenhoven et al (2007) speculated th at differen­ ces in ANL growth (i.e., the amount th at ANLs change with increases in speech presentation level) may be caused by differences in loudness recruitment between normal-hearing and hearing-impaired listeners. How­ ever, the results of their study showed that ANL growth was not related to listeners’ hearing sensitivity. In another study, Freyaldenhoven et al (2008b) sug­ gest th at unsuccessful hearing-aid wearers may reject hearing aids because of a few listening situations in which they do not accept background noise. If this is true, testing ANLs at multiple levels may help predict who will wear hearing aids full time, part time, or not at all. However, the results of their study, which examined a 35-dB test range (speech fixed at 40-75 dB HL), found th at ANL growth predicted hearing-aid use slightly less well than ANL scores by themselves (ANLs predicted hearing-aid use with 68% accuracy, and ANL growth predicted hearing-aid use with ^63% accuracy). Mild correlations have been found between listeners’ ANLs and their perceived concentration levels, and their perceived speech understanding abilities in background noise (Recker et al, 2011). Additionally, for a small subset of individuals, ANLs equal minimal acceptable speech levels (MinASLs) (Recker and Edwards, 2013). MinASLs are defined as the difference between the minimal level at which someone is willing to listen to speech and the level of the background noise. In that study, a range of BNLs was used, from 50 and 80 dBA, to represent the range of levels th at are encountered in everyday listening environments. To date, mixed results have been found regarding ANLs and an individual’s ability to understand speech in noise. Crowley and Nabelek (1996) and Nabelek et al (2004) found that the two measures are uncorrelated, whereas Ahlstrom et al (2009) found that people with better speech intelligibility have lower ANLs. Addition­ ally, it is currently unclear whether ANLs are affected by the use of amplification. Nabelek et al (2004) found no difference between ANLs th at were obtained aided and unaided, whereas Ahlstrom et al (2009) found that

A c c e p ta b le N o ise L e v e ls a n d L o u d n ess/R eck er et al

people have lower ANLs when they are tested aided than when they are tested unaided. If it were known w hat cues listeners are using to decide how much background noise they are willing to tolerate, then it might be possible to create technology that addresses these cues and therefore improves listen­ ers’ chances of success with hearing aids. In fact, several studies have already shown that certain hearing-aid fea­ tures (e.g., directional microphones) and types of signal processing (e.g., noise reduction algorithms) can increase the amount of background noise that a listener is willing to tolerate by approximately 2.5-4 dB (Freyaldenhoven et al, 2005; Mueller et al, 2006; Peeters et al, 2009). As a first step in determining why listeners are will­ ing to tolerate the amount of background noise th at they are willing to tolerate, this study investigated whether listeners’ ANLs are consistent with a loudness-based listening strategy. Loudness was chosen because it is easy to conceive th at people who are not willing to put up with high levels of background noise are doing so because of loudness-tolerance issues. That is, listen­ ers are willing to tolerate background noise until the point th at it becomes “too loud.” Once this point is reached, they are not willing to tolerate increases in the loudness of the background noise. If this were true, one would expect the BNL, in the ANL test, to equal the listener’s maximal tolerable loudness for that sound. Furthermore, one would expect the maximal tolerable loudness of the background noise to remain constant even as the presentation level of the speech changes. This is not to say th at the level of the background noise would rem ain constant as the level of the speech changes. When two sounds are presented at the same time, as occurs with the ANL test, the presence of one sound alters one’s judgment of the loudness of the other sound, a concept called “partial loudness” (Moore et al, 1997). In general, when two sounds are presented at the same time, if the cochlear excitation of one stimulus is much higher than the cochlear excitation of the other stimulus, the loudness of the stimulus eliciting the higher cochlear excitation will approach the loudness of that stimulus in quiet, and the loudness of the stimulus eliciting the lower cochlear excitation will approach zero. Additionally, when one stimulus is at its masked threshold (i.e., the threshold for detection in the presence of the other stim­ ulus), its partial loudness will equal the loudness of that stimulus in quiet at the absolute threshold (i.e., the loud­ ness will be quite low). It is important to note that having a second stimulus present will never increase the loudness of the first stimulus; it can only decrease it. As a simple example of partial loudness, if someone were yelling at 80 dBA in a quiet room, his or her voice would seem quite loud. However, if this same voice were yelling in a crowded restaurant, where the background noise was also at 80 dBA, it would sound much less loud.

If a listener were asked to match the loudness of the two voices, he or she would either need to increase the level of the person yelling in the restaurant or decrease the level of the person yelling in the quiet room. In fact, this is exactly how partial loudness has been studied by researchers (Gassier, 1954; Zwicker, 1963; Scharf, 1964; Stevens and Guirao, 1967; Heilman, 1970; Houtgast, 1974; Spiegel, 1981; Gockel et al, 2003). In these studies, participants heard both a reference and a test stimulus. The reference stimulus included a target sound presented in quiet, and the test stimulus included both a target and a masker. Participants adjusted the level of the target in one stim ulus until it was equally loud as the target in the other stimulus. In some studies, the level of the target in the reference, stim ulus was adjusted; in other studies, the level of the target in the test stimulus was adjusted. To date, almost all partial-loudness studies have used simple stimuli. Tones, complex tones, and narrowband noises have been used as targets and narrowband and broadband noise have been used as maskers. Only one study could be found in the literature that used a more real-world stimulus. In 1949, Pollack used white noise as a masker and continuous recorded speech (Adam Smith’s Wealth o f Nations) as the target. There were only two participants in his study. Both participants adjusted the level of the speech, in quiet, to match the level of the speech presented in white noise. Results of that study showed little change in the perceived loud­ ness of the speech until the noise was within 5 dB of the level of the speech. As the signal-to-noise ratio (SNR) became poorer, the loudness of the speech decreased rapidly. Additionally, Pollack found th at the effect of the noise on the loudness of the speech was a function of the SNR rather than the absolute level of the speech or the white noise. Because the loudness of a sound will vary depending on what other sounds are present in the environment, in order to determine whether listeners are using loudness as a cue to set the level of the background noise that they were willing to accept for the ANL test, it is first necessary to understand how the presence of the speech s ig n a l affects listeners’ judgments of the loudness of the back­ ground noise. Because the loudness of the background noise will vary depending on the level of the speech signal, multiple levels will be tested. In summary, the goal of this study is to determine whether the noise levels that participants choose for the ANL test are consistent with a loudness-based listen­ ing strategy. This will be accomplished by testing listeners ANLs at several different fixed speech levels and compar­ ing the BNLs that they choose with their equal-loudness contours for the background noise. Traditionally, equal­ loudness contours have been used to represent the levels at which pure tones of different frequencies are considered to be equally loud; in this context, we are using the term to

607

Journal of the American Academy of Audiology/Volume 25, Number 6, 2014

represent the levels at which multitalker babble, in the presence of speech presented at different levels, are con­ sidered to be equally loud. For ANLs that run parallel to the equal-loudness contours, the BNLs will be considered to be equally loud and therefore consistent with a loudnessbased listening strategy.

METHODS Equipment For all testing, the speech was the Arizona Travelogue (Cosmos Dist. Inc., nd) and the noise was multitalker babble. Although no information was provided with the Cosmos compact disk (CD) regarding the multitalker babble, other researchers (e.g., Nabelek et al, 2006) have referenced this as 12-talker speech babble (revised SPIN recorded by Cosmos Dist. Inc., nd; Bilger et al, 1984). Both of these stimuli were extracted from the CD and saved as .wav files on a local hard drive for use during an automated version of the ANL test. (Unless otherwise stated, for the remainder of this article, it should be assumed that any references to “speech,” “noise,” or “background noise” for this study are references to the Arizona Travelogue and multitalker babble.) Audio stimuli were routed from a Dell Precision T3500 computer to a RME Multiface II (audio interface) to an AMX Autopatch Precis DPS to a Tannoy 6D loud­ speaker. The test was calibrated by playing a calibra­ tion noise through the loudspeaker and adjusting the level until it was 65 dBA1 at the location of the partici­ pant’s head, 3 feet from the loudspeaker at a 0° azimuth. (Because the Cosmos CD did not include a calibration stimulus, this was created by shaping a white noise to match the long-term spectrum of the speech and multi­ talker babble stimuli.) During the study, participants adjusted the levels and marked their responses using custom graphic user interfaces (created using Max/MSP) and a MultiTouch LCD 175VXM touchscreen monitor. The video signal was routed from the Dell computer to a Kramer VS-66HDCP to the touchscreen monitor. All testing was performed in a sound-treated booth.

Participants A total of 21 normal-hearing and 21 hearing-impaired individuals participated in this study. Most of these individuals were recruited from Starkey’s internal research database. This database includes the names of hundreds of individuals who have volunteered to par­ ticipate in research with Starkey. Normal-hearing indi­ viduals were primarily Starkey employees who have responded to companywide e-mails seeking volunteers 1 If the reader wishes to convert the dBA values in this article to dB SPL (no weighting), 6 dB should be added.

608

to participate in research. Hearing-impaired individuals were primarily individuals who have responded to recruit­ ment advertisements that were placed in local newspa­ pers. Participants were selected based on audiometric criteria and ANL testing with the goal of obtaining seven participants, both normal-hearing and hearing-impaired, from each of the ANL groups (low, mid, and high). For normal-hearing participants, it was desired that all audiometric thresholds equal 20 dB HL or less and th at the thresholds for the two ears be symmetric (within 10 dB at each frequency). However, because of the difficulty in finding participants who also met the ANL criterion, as will be discussed later in this sec­ tion, the audiometric criteria were relaxed slightly. Specifically, for octave frequencies 250-8000 Hz, five of the normal-hearing participants had one or more thresholds of 25 or 30 dB HL, and two individuals had asymmetry of 15 dB at 1 frequency. All of the individuals who had audiometric thresholds a t 25 or 30 dB HL had traditional ANLs that placed them in the mid-ANL group (7-13 dB) or the high-ANL group (>13 dB) (“tradi­ tional ANL” meaning the test followed the original for­ mat in which individuals first adjusted the speech to their MCLs, and then background noise was added and they adjusted it to the maximal levels that they were willing to tolerate while listening to the speech). One individual from the low ANL group and one individual from the high ANL group had an asymmetric threshold. For the hearing-impaired participants, it was desired th at all participants have symmetric sensorineural hear­ ing loss in the mild-to-moderate range. However, as with the normal-hearing individuals, because of the difficulty in finding participants who also met the ANL criterion, the audiometric criteria were relaxed. Fifteen of the 21 participants had asymmetry of 15 dB or more at one or more of the octave frequencies between 250 and 8000 Hz. Below 4000 Hz, nine individuals did not have any asym­ metry in their hearing thresholds, nine individuals had asymmetry of 15 dB at a single frequency, two individ­ uals had asymmetry of up to 20 dB at 2 frequencies and one individual had asymmetry of 25 dB at 1 fre­ quency. Larger hearing losses were avoided in this study because all of the testing was performed unaided. Figure 1 shows the mean hearing thresholds ± 1 SD for the low, mid, and high ANL groups (both normal­ hearing and hearing-impaired). Hearing thresholds were not matched across ANL groups for this study because Nabelek et al (1991) suggested that ANLs are not related to an individual’s hearing sensitivity. For this study, it was desired th at an equal number of individuals participate from the three ANL groups (low, mid, and high). Therefore, each study candidate was screened with an automated version of the traditional ANL test. Specifically, participants were instructed to follow these steps: Please adjust the volume of the speech to a level that is...

A cceptable N oise Levels and Loudness/Recker et al

Figure 1. Mean and range of audiometric thresholds for study

participants.

1. Too loud 2. Too soft 3. Most comfortable to you Next, multitalker babble was added, and participants were instructed to: Please adjust the volume of the background noise to a level that is... 1. Too loud to understand the speech 2. Soft enough for the speech to be very clear 3. The maximal noise level that you would be willing to “put up with” for a long time while following the story The first time each person performed the ANL test, the speech and the noise were presented at 35 dBA. For all subsequent testing, the initial presentation level of the speech and the noise was random, between 35 and 59 dBA. For instructions 1 and 2, the step size was 5 dB; for instruction 3, the step size was 2 dB. The final instruction was the one of interest; the purpose of the initial instructions was to get participants to explore the range of levels that was available to them. Verbal instructions were given before the testing began, and written instructions appeared on the participants’ monitor during the testing. The ANL = MCL - BNL. Individuals performed the traditional ANL test five times. The first iteration was practice; iterations 2-5 were averaged to determine the individual’s ANL. Because of the difficulty in finding listeners with hearing loss in Starkey’s research database with ANLs greater than 13 dB, an advertisement was placed in a local newspaper to recruit additional participants. The advertisement was specifically aimed at recruiting individuals who have been unsuccessful with hearing aids in the past. These individuals were targeted because Nabelek et al (2006) suggested that people with high ANLs are unlikely to be successful with hearing aids; it was hoped that the opposite was also true - that people who have been unsuccessful with hearing aids also have high ANLs.

In all, 79 normal-hearing and 141 hearing-impaired people were screened with the ANL test to find partici­ pants for this study. Of the normal-hearing individuals, 67% had a low ANL (13 dB). Of the hearingimpaired individuals, 53% had a low ANL, 35% had a mid ANL, and 11% had a high ANL; one additional individual’s results were too inconsistent to categorize his score (it ranged from 1.8-45 dB). Table 1 summa­ rizes the MCLs, BNLs, and ANLs (means, SDs, and ranges) for the study participants and for the entire sample that was screened with the ANL test. Appro­ ximately half of the individuals with hearing loss who were screened with the ANL test and who had high ANLs did not qualify for this study because they had hearing losses that were too severe for unaided testing, they had hearing losses th at were asymmet­ ric, or they had difficulty understanding the test instructions. In addition to ANL and audiometric testing, most of the individuals who were screened for this study took a short questionnaire that investigated the negative impact that background noise had on their perceived speech intelligibility, stress levels, and concentration levels. Furthermore, the questionnaire asked individ­ uals about their own perceived tolerance for background noise and whether they avoid situations known to have high levels of background noise. Finally, it asked indi­ viduals to describe their hearing-aid use following the categories that were defined by Nabelek et al (2006): I wear my hearing aids whenever I need them, I only wear my hearing aids occasionally, or I do not wear my hearing aids. Although the results of this question­ naire have already been published (Recker et al, 2011) and will not be discussed in detail in this article, it is worthwhile to mention the results of this final question for the individuals who participated in this study. Of the 21 hearing-impaired individuals who participated in this study, all but 5 of them (2 individuals in the low-ANL group, 1 individual in the mid-ANL group, and 2 individuals in the high-ANL group) reported that they owned hearing aids. One additional participant, in the low-ANL group, reported that he owned hearing aids but only wore them occasionally (oS tl- P- p- COC\J 00 oo re CDCDLOq ~D re II CDre d cd pd LOo cd COre re 1 CO c CD LOp- re re LO 1

Q) Q.

E

(0 (/) o )k— ■3

T3 0

S C M C M C O C O ^ C \ j C \ j l f i S r ;

oc\j^j-T--T-OLOcjjT-r^

c

(OT-^COlfir-CON

o

LU

d) £ O O C D C O ’- L O l O N C O C D N C O

cocod'tibd^Nc\i H ^CD( D ci n d CD lO r N r 1-

f i

TJ C ■D 3

co 1 c

75

u 0 c 0 0 b CO

re re re q q CD CO q LO T- q CD q CM cd re CD pd cd 'sf q LO d CVJ CD LO CO oo LO 1

_i

>> , pTO =5 re II cd CL c

q ^p T- CM q q re CO CD co 00 cd CM re T— re cd co d re1 T— T— CD LO CD LO ■^p LO LO

0 _l z
c

E*a oc -§

< > _l

g 8 o g

° 8 £ 3 03

co

co

in

co

CC

>,

Q. X LU

.9

I

o c c

0 0 0 >


8 © S 3

O)

03

“ « « c r co go

I CO

U ) 03

O >

>

-X

I f K

(1)

r 03

>,

-££

CM

o

(/) D

CL

c

O

“ §v-

CO

CO CM CC

(0

X

o

CO

CO 0

_0

CM ■«—

c 'w (0 0

CO

£

0 t= CO 03

O CM

LU g o a E cc cc t

a> o > 0 a TJ C


. 0 JxC 0 CO

o

X

Q to o h*. 0 ■sr Q

0 >* .9

X

CO


. _c

Note: Participants are ordered according to their traditional ANLs.

£

o o

6 _l

•iz 0 0 X

£1 LL

0 ^ CO 03 CO £

to 0 Q

8 $ Q
70

.

"TfiB-"1*

reference noise 55(dBA) 50

60

70

Speech Level (dBA)

60

80

40

50

60

70

80

90

Speech Level (dBA)

70

Speech Level (dBA)

Figure 4. Loudness-matched results for multitalker babble in the presence of speech for normal-hearing (left) and hearing-impaired (right) participants. The level of the reference condition was altered in 10 dB steps from 85 dBA (top row) to 45 dBA (bottom row).

A c c e p ta b le N o ise L e v e ls a n d L o u d n ess/R eck e r et al

loud and therefore consistent with a loudness-based listening strategy. ANLs th a t did not run parallel with the equal-loudness contours were not considered to be equally loud and therefore were not consistent with a loudness-based listening strategy. As an additional task, participants performed the loudness-m atching task without the speech signal present. This was done to determine how well the listeners could perform the loudness-matching task in the absence of the speech stimulus. For the loudness-matching task, the reference and the test stimulus both played for 3 sec. Both samples were chosen randomly from the .wav files. The refer­ ence played first and then the test condition. After each stimulus played once, participants were permitted to toggle back and forth between the test and the reference stimuli as many times as desired before confirming their response. The level of the background noise of the test stimulus was changed by moving a slider on a touchscreen monitor. All loudness-matching conditions were repeated a t least three times and were averaged. If there was greater than 6 dB of variability across the participant’s three responses, additional trials were added until the median three responses (of all of those th a t were per­ formed) were within 6 dB. W ritten and verbal instruc­ tions were provided to the participants before testing began (see Appendix A for a copy of the participant’s instructions). ANL testing was completed during one session and loudness-matching testing during a second session. All testing was performed unaided.

RESULTS he goal of this study was to determine whether the maximum noise levels th at listeners were willing to accept for the ANL test were consistent with a loudnessbased listening strategy. To do this, each participant com­ pleted ANL and loudness-matching testing at multiple levels.

T

ANLs Figure 2 shows a summary of the ANLs for normal­ hearing (left) and hearing-impaired participants (right). Participants’ ANL results are grouped based on each individual’s traditional ANL (i.e., whether it fell into the low [13 dB] ANL group). For both the normal-hearing and the hearingimpaired groups, there was an approximate 20-35 dB range of SNRs that was considered acceptable for the different speech presentation levels. For most participants, as the level of the speech increased, ANL also increased. This concept is known as “ANL growth,” and it is defined as the change in ANL divided by the change in speech presentation level (Franklin et al, 2006; Tampas and Harkrider, 2006; Freyaldenhoven et al, 2007, 2008b). For the normal­ hearing participants, ANL growth was 0.11, 0.17, and 0.17 dB/dB for the low-, mid-, and high-ANL groups, respectively, and 0.08, 0.14, and 0.19 dB/dB for the hearing-impaired low-, mid-, and high-ANL groups, respectively. Table 3 provides a summary of the individ­ ual ANL growth rates, along with the mean and SD for each group. To determine whether there were any significant dif­ ferences in ANL growth among the different groups, a two-way analysis of variance was completed with hearing status (normal or impaired) and ANL group (low, mid, or high) as factors. Results showed no significant differences in the results based on hearing status [F(l,36) = 0.0620, p = 0.805)] or ANL group LF(2,36) = 1.226, p = 0.306]. Additionally, there was not a significant interaction between these two factors [F(2,36 = 0.124, p = 0.884)].

Loudness Matching To determine how well participants were able to perform the loudness-matching task, as part of the testing, parti­ cipants performed the loudness-matching task without

T a b le 4. S u m m a r y o f th e R e s u lts o f th e H o lm -S id a k A ll-P a ir w is e C o m p a ris o n S ta tis tic C o m p a rin g P a r tic ip a n ts ’ N o is e M a tc h e d L e v e ls fo r E a c h o f th e D iffe re n t R e fe r e n c e C o n d itio n s a n d S p e e c h P r e s e n ta tio n L e v e ls

p Values for Each Reference Noise Level (dBA) Normal Hearing Hearing Impaired Comparison Speech Levels (dBA)

88 88 88 75 75 63

vs. vs. vs. vs. vs. vs.

50 63 75 50 63 50

Normal Hearing and Hearing Impaired (Results Collapsed)

45

45

55

65

75

85

Loudness as a cue for acceptable noise levels.

The acceptable noise level (ANL) test is the only test that is known to predict success with hearing aids with a high degree of accuracy. A person's A...
12MB Sizes 0 Downloads 9 Views