Establishment of Age-Specific Normative Data for the Canadian French Version of the Hearing in Noise Test for Children Ve´ronique Vaillancourt,1 Chantal Laroche,1 Christian Gigue`re,1 and Sigfrid D. Soli2

than 0.5 dB for the noise conditions but exceeded 4 dB in the quiet condition. The reliability of SRT measures was determined, with an overall withinsubjects SD of repeated measurements of 0.7 dB for the noise front condition. No learning effect was found in the current data.

Objectives: A Canadian French version of the Hearing in Noise Test (HINT) has been developed to assess children’s ability to recognize speech in noise. To avoid testing a large number of children in each clinical test site to establish soundfield norms, a protocol based on the use of correction factors has been proposed and validated in the current study. More specifically, the objective of this study was to provide a protocol for the establishment of agespecific normative data for the Canadian French HINT for children to facilitate its clinical use and allow comparing an individual child’s performance with that of age-matched normal hearing children. Using the proposed protocol, a limited number of normal hearing adults are tested in each HINT condition to correct the adult headphone norms for the soundfield in question, and the correction factors established in the current study are then applied to generate age-specific soundfield norms. Mean adult performance values obtained in a given soundfield are entered into the HINT software, which automatically derives the soundfield adult norms, age-specific children norms, and percentile rankings.

Conclusions: Correction factors can be used to predict performance on the HINT in a group of normalhearing children in all HINT conditions, apart from quiet. Findings of the current study concur with the literature on age effects in auditory processing abilities, where performance on a variety of auditory tasks has been demonstrated to increase with age to reach adult-like values in adolescence or past 10 yrs. (Ear & Hearing 2008;29;453– 466)

INTRODUCTION Speech understanding, especially in noise, can be significantly hindered by hearing loss, as demonstrated by the frequent accounts of difficulties hearing in noise among individuals with hearing loss (Gatehouse & Noble, 2004) and hearing aid users (Kochkin, 2000, 2002). Although other hearing difficulties also strongly contribute to handicap (Gatehouse & Noble, 2004), it is generally accepted that listeners with sensorineural hearing losses typically require more favorable signal to noise ratios (SNR) to understand speech at a performance level similar to that achieved by normally hearing listeners (Bronkhorst & Plomp, 1990; Dubno, et al., 1984; Gelfand, et al., 1988; Plomp, 1977; Rowland, et al., 1985). Moreover, individuals with similar configurations and degrees of hearing impairment differ significantly in their ability to understand speech in noise (Crandell, 1991; Killion & Niquette, 2000; Nabelek & Pickett, 1974). Indeed, previous research clearly indicates weak predictive correlations between puretone thresholds and self-reports, questionnaires, and measures of speech recognition in noise (Bronkhorst & Plomp, 1992; Duquesnoy, 1983; Plomp & Mimpen, 1979; Smoorenburg, 1992). Consequently, traditional hearing assessments based on pure-tone threshold measurements cannot reliably quantify the ability to understand speech in noise or predict other hearing

Design: Speech reception thresholds (SRT) for sentences were measured in 70 native French-speaking subjects to establish mean performances across various age groups, and correction factors were calculated by comparing performance in each age group with adult performance. To validate the normalization protocol, 28 additional subjects were tested in a new soundfield. The correction factors were applied to adult performance (N ⴝ 15) and the resulting predicted scores were compared with measured performance in a group of 9-yr olds (N ⴝ 13). Results: Statistical analyses indicate that SRTs decrease with age and reach adult values in older children (12-yr olds). Correction factors are therefore provided for children 6 to 12 yrs old. Spatial separation advantage, the improvement in SRT when speech and noise are spatially separated, also improves with age. The correction factors were effective in predicting mean SRTs for a previously untested age group in all HINT conditions apart from the quiet condition. The difference between predicted and measured performances was less 1

Audiology and Speech-Language Pathology Program, Faculty of Health Sciences, University of Ottawa, Ottawa, Ontario, Canada; and 2Human Communication Sciences and Devices Department, House Ear Institute, Los Angeles, California.

0196/0202/08/2903-0453/0 • Ear & Hearing • Copyright © 2008 by Lippincott Williams & Wilkins • Printed in the U.S.A. 453

454

VAILLANCOURT

ET AL.

/ EAR & HEARING, VOL. 29, NO. 3, 453–466

difficulties encountered by individuals with hearing impairment (Killion & Niquette, 2000). In the hope of better quantifying the ability to understand speech, many have developed and used sentence-length materials to perform speech recognition testing in various languages (e.g., Bench, et al., 1979; Cox, et al., 1987; Hagerman, 1982; Kalikow, et al., 1977; Killion & Villchur, 1993; MacLoed & Summerfield, 1990; Plomp & Mimpen, 1979). Despite the fact that their use in clinical settings remains rather limited (Medwetsky, et al., 1999; Mueller, 2001, 2003; Strom, 2003), sentence-length materials are more representative of everyday communication than isolated words (Nilsson, et al., 1994) and are therefore better suited for use in a comprehensive evaluation of functional hearing abilities (Bosman, 1989) or to predict one’s performance in everyday or workplace communication situations (Laroche, et al., 2003). Shortcomings in a number of sentence-length speech tests available in American English [e.g., Connected Speech Test (Cox, et al., 1987) and Speech Perception in Noise Test (Kalikow, et al., 1977)] have led Nilsson et al. (1994) to develop the Hearing in Noise Test (HINT), founded on the earlier work of Plomp and Mimpen (1979). Indeed, floor and ceiling effects typically plague measures performed at fixed presentation levels (MacLeod & Summerfield, 1987; Nilsson, et al., 1994) and scoring based on the correct repetition of single words (e.g., the final word) can reduce efficiency in clinical settings since much of the material within a sentence is not scored. Adaptive speech tests and scoring of multiple words or entire sentences have been proposed to overcome these limitations. The HINT (Nilsson, et al., 1994) exploits an adaptive method and scoring based on the correct repetition of entire sentences to determine the speech reception threshold (SRT) for sentences, defined as the presentation level at which a listener can correctly repeat the sentences 50% of the time. Functional hearing is assessed under headphones or in a sound field by measuring SRTs in quiet and in three conditions of speech-spectrum noise: (a) noise front (NF, speech and noise in front at 0° azimuth), (b) noise right (NR, speech in front and noise at 90° azimuth), and (c) noise left (NL, speech in front and noise at 270° azimuth). The spatial separation advantage can also be obtained by subtracting noise side (NS) score from NF. This measure, ranging from 6 to 10 dB in normally hearing adults, highlights the advantage associated with the spatial separation of speech and noise sources for understanding speech in noise (Bronkhorst & Plomp, 1988; Soli & Nilsson, 1994). Speech understanding in children is similarly important. Like adults, children with hearing loss require more favorable SNRs than normally hearing

children for equal speech recognition performance (Ruscetta, et al., 2005). In addition, some children are faced with auditory processing disorders (Chermak & Musiek, 1997). Difficulties hearing in noise are a common occurrence and can have profound consequences on learning (Bellis, 2003; Picard, 2003; Picard & Bradley, 2001). Therefore, child versions of the HINT (HINT-C) have been developed in English (Nilsson, et al., 1996) and in Canadian French (Laroche, et al., 2006) to help quantify these hearing difficulties. The Canadian French HINT-C (Laroche, et al., 2006) consists of a fixed 65-dBA speech-spectrum masking noise and 170 short sentences (5–7 syllables) extracted from the 240 sentences of the adult version (Vaillancourt, et al., 2005), which were derived from the vocabulary corpus of 6-yr-old children (Leduc, 1997). Only those sentences with a recognition rate greater than 94% (mean ⫽ 99%, SD ⫽ 1.5%) when presented to a group of 5-yr-old children, and which did not include words that were unfamiliar or difficult to pronounce, were included in the final speech set for children. Such criteria ensured that the material would be suitable for children 6 yrs and older. These sentences, equated for intelligibility, were then grouped into 17 phonemically balanced 10-sentence lists. SRTs are measured in each HINT condition (Quiet, NF, NR, and NL) using 10-sentence lists. Using the Canadian French HINT-C material, Laroche et al. (2006) gathered preliminary data on HINT performances as a function of age in 58 school-aged children [6 (N ⫽ 14), 7 (N ⫽ 15), 8 (N ⫽ 14), and 9 (N ⫽ 15) yr olds] in a Noise Front condition. By means of the standard protocol proposed by Nilsson et al. (1996) for testing children, a SRT was calculated for each of three 10-sentence lists and the two best scores were averaged to obtain a final performance score (SRT) for each child. This approach controls for potential learning effects and moments of inattention during testing. Using this procedure, the reliability of SRT measurement, calculated as the within-subjects SD of repeated measurements, was found to be similar (about 1 dB) to that obtained with the adult French version (one 20-sentence list per condition) (Vaillancourt, et al., 2005), and both the adult and children English versions (Nilsson, et al., 1994, 1996). Furthermore, mean group SRTs were shown to be statistically significantly different at an alpha level of 0.05 for age groups separated by at least 2 yrs (6 – 8, 6 –9, and 7–9), but not for successive age groups (6 –7, 7– 8, and 8 –9). Headphone administration of the HINT entails processing the signals with digital filters recreating the head-related transfer functions (KEMAR HRTFs) to simulate the four free-field conditions: quiet, NF,

VAILLANCOURT

ET AL.

/ EAR & HEARING, VOL. 29, NO. 3, 453–466

NR, and NL (Soli & Nilsson, 1994). Because children and adult HRTFs differ significantly (Fels & Vorla¨nder, 2004) and age-specific HRTFs are thus far not available to accurately reproduce the interaural differences in signals reaching both ears, soundfield administration is the only reliable method for testing children. To conduct soundfield testing, two loudspeakers forming a 90° angle are placed 1 m from the center of the listener’s head. Because HINT scores are somewhat dependent on room acoustics (reverberation time, reflections, objects acting as obstacles to sound propagation, etc.) (Lamothe, et al., 2002; Nilsson, et al., 1996), roomspecific normative values allowing the comparison of a child’s performance with that of age-matched normal-hearing children must be established for each site where testing occurs. Given age-related differences in performance (Laroche, et al., 2006; Nilsson, et al., 1996), a traditional protocol for establishing normative values would be clinically unpractical and unfeasible, requiring testing a large number of children of various ages in each clinical test site to establish site-specific normative values for each age group. Nilsson et al. (1996) proposed a practical alternative whereby a single set of age-specific correction factors relative to adult performance are used to establish normative values. Using their approach, schematized in Figure 1, adult performance is measured in a given soundfield to account for the specific acoustical characteristics of the clinical setting. The age-specific correction factors are thereafter applied to these adult soundfield norms to generate age-specific normative values for the soundfield in question. The objective of this study was to establish correction factors for the Canadian French HINT-C, using the approach proposed by Nilsson et al. (1996), and validate their use in a different soundfield. More specifically, the objectives of this study were to (a) measure performance in a given soundfield for normal hearing children of various age groups (6, 8, 10, and 12 yrs) and normal-hearing adults, to establish correction factors that can ultimately be applied

Fig. 1. Approach proposed by Nilsson et al. (1996) to generate age-specific norms and percentile rankings for children.

455

to generate normative data in multiple settings (experiment 1), and (b) validate this approach for a group of 9-yr-old children in a soundfield different than that used in experiment 1, by comparing measured and predicted mean performance scores for this group (experiment 2).

MATERIALS

AND

METHODS

Experiment 1—Age-Specific Correction Factors for Normal Hearing Children Subjects • A total of 70 native French-speaking subjects with normal hearing who speak French on a daily basis took part in this study, including 14 children in each of four different age groups (6, 8, 10, and 12) and 14 adults between the ages of 18 and 30 (mean age ⫽ 24). The age groups used in the current study were chosen on the basis of preliminary data using the HINT-C obtained by Laroche et al. (2006), in which statistically significant differences in mean SRTs were found only for groups separated by at least 2 yrs, but not for successive age groups. It was therefore proposed to use groups for which significant differences in SRT were likely to occur. Sample size was also chosen based on a priori knowledge and statistical power estimates. To motivate establishing an age trend rather than grouping data from all children, performance in the two extreme age groups of children must be different. SDs and differences in mean predicted SRTs between children 6 and 12 yrs old from the English HINT-C (Nilsson, et al. 1996) were used to estimate the difference in units of SD (⌬) between the two extreme groups under study, where ⌬ corresponds to the difference in SRT divided by the SD. For the quiet, NF, and NS conditions, ⌬ values of 3.1, 1.4, and 2.2 were obtained, respectively. Using the smallest ⌬ value (1.4), ␣ ⫽ 0.05 for a single-sided test, and the Student t distribution—number of observations of t test of difference between two means (Duncan, et al., 1983), 8, 10, 12, and 17 subjects are required in each group for ␤ values of 0.2, 0.1, 0.05, and 0.01, respectively. Fourteen subjects was therefore deemed an appropriate choice in sample size, especially since statistically significant differences between age groups separated by at least 2 yrs had previously been found by Laroche et al. (2006) with 14 to 15 subjects per group. Furthermore, as a difference smaller than 1 SD between any two groups would lack clinical significance, it can be stated that using the proposed sample size (N ⫽ 14), one can expect to detect, with 80% certainty (␤ ⫽ 0.2), a difference as small as 1 SD (⌬ ⫽ 1), if it exists, in the mean SRTs when ␣ ⫽ 0.05 in a single-sided test.

456

VAILLANCOURT

ET AL.

/ EAR & HEARING, VOL. 29, NO. 3, 453–466

Subjects were recruited from two typical French public schools in Ottawa and Gatineau, and were educated in French and spoke French on a daily basis. Only native speakers were selected, because age of language acquisition and experience with language are known to influence speech recognition performances (van Wijngaarden, 2003). Normal hearing was defined as pure-tone air conduction thresholds equal to, or better than 15 dB HL (ANSI S3.6 –1996) from 250 to 8000 Hz, bilaterally. Subjects had normal otoscopic examination, normal tympanograms, and a negative otologic history. Moreover, they were free of suspected or diagnosed language difficulties, developmental problems, learning disabilities, and difficulties in school, as reported by their parents and teachers. They were therefore considered to have normal language development, although this was not specifically assessed through formal language inventories. Procedure • Before testing, the subjects (or parents) were required to read an information letter, sign a consent form, and fill out a hearing and developmental history questionnaire. The hearing screening was performed in the same sound booth as the HINT testing, using pure tones, the HughstonWestlake up– down technique (Hall & Mueller, 1997), and a portable audiometer (InterAcoustics AD25) with TDH-39P headphones. A WelshAllyn otoscope and GS1738 impedance bridge were used to assess the integrity of outer and middle ears. Administration of the HINT-C was achieved with a desktop computer coupled with the HINT hardware (hearing test device, talk-back microphone, talk-forward microphone, USB cable, power supply, and tester’s headset), the HINT for Windows software, and two Optimus XTS40 speakers. The hearing test device serves as an external sound card to which all the components are connected. The speakers were placed 90° apart in an IAC audiometric room (2.3 ⫻ 2.2 m2), 1 m away from the center of the listeners’ head. The testing apparatus was calibrated by following the on-screen step-by-step calibration procedures in the HINT software and described in the user’s manual. While some steps are automatically performed by the software, such as verification of speaker matching, the user must enter measured values in the software when performing calibration for sound levels. Calibration was performed using a B&K2235 sound level meter with B&K1625 filter set, a B&K4176 microphone, and a B&K4228 sound calibrator. Soundfield SRTs were measured in each HINT condition (quiet, NF, NR, and NL) using 10-sentence lists, scoring on whole sentences, and the standard default parameters (noise level, starting level or SNR, and step sizes). The subjects were instructed

to repeat everything they heard, and to guess if necessary. A 10-sentence list was initially presented in quiet to familiarize the subject. The initial sentence in a list was presented at 20 dBA in the quiet condition, at 0 dB SNR in the NF condition, and at ⫺5 dB SNR in the NR and NL conditions. Speech levels were reduced automatically by the software following a correct sentence repetition and increased following an incorrect response, whereas the noise level was fixed at 65 dBA. Steps of 4 dB were used to adjust the presentation level of the first four sentences, whereas 2 dB steps were used thereafter. The SRT is estimated as the average level or SNR of sentences 5 to 11 (the 11th sentence in not presented, but its level is known from the previous response). Using the standard protocol proposed by Nilsson et al. (1996) for testing children, the SRT was calculated for each of three 10-sentence lists in a given condition and the two best scores were averaged to obtain a final performance score (SRT) for each child. Testing began with a practice list to familiarize the subjects, followed by SRT measurement in quiet. For the noise conditions, testing order was counterbalanced between subjects. The HINT software randomly chose both list order for each test condition and sentence order within each list. A short break followed each HINT condition to limit the effect of fatigue on performance and maintain subject motivation. Results Mean Performances • For each subject, an overall score was calculated based on two measurements for each HINT condition, as discussed earlier, and served as the dependent variable in the statistical analyses. All analyses were performed at a significance level of ␣ ⫽ 0.05 using SPSS version 12.0. A mixed design analysis of variance (ANOVA) with one independent repeated-measures variable (HINT condition) and one between-subjects variable (age) showed a significant effect of HINT condition [F(3, 222) ⫽ 9441.170, p ⬍ 0.001], a significant effect of age on SRT [F(5, 74) ⫽ 109.878, p ⬍ 0.001], and a significant interaction of HINT condition and age group [F(15, 222) ⫽ 15.119, p ⬍ 0.001]. As withinsubjects contrasts did not show a significant difference between the NR and NL conditions [F(1,74) ⫽ 0.344, p ⫽ 0.559], or a significant difference in the interaction between HINT condition and age group for NR and NL [F(5,74) ⫽ 1.762, p ⫽ 0.131], the overall NR and NL scores were averaged to yield an NS score. Moreover, the spatial separation advantage (or advantage for spatial separation of speech and noise) was computed by subtracting NS from NF. Group mean SRTs and spatial separation advantage are displayed in Figures 2 and 3, respectively.

VAILLANCOURT

ET AL.

/ EAR & HEARING, VOL. 29, NO. 3, 453–466

457

Fig. 3. Mean spatial separation advantage as a function of age. SD error bars are also shown (ⴞ1.96 SD). Adult data are included for comparative purposes.

measure by means of a t test for multiple comparisons using a Bonferroni adjustment (Duncan, et al., 1983). Results are summarized in Table 1. Briefly stated, most differences between groups reached statistical significance, except for the following: (a) quiet: 8 to 10 and 12-adults, (b) NF: 8 to 10, 8 to 12, 10 to 12, and 12-adults, (c) NS: 12-adults, and (d) spatial separation advantage: 6 to 8, 8 to 10, and 12-adults. Although successive age groups were not TABLE 1. Statistically significant differences at a 0.05 level of confidence, performed by means of a t test for multiple comparisons using a Bonferroni adjustment

Measure Quiet

Noise front

Fig. 2. Mean speech reception thresholds for sentences (SRT) as a function of age for each HINT condition (quiet, noise front, and noise side). SD error bars are also shown (ⴞ1.96 SD). Adult data are included for comparative purposes. Noise side

Typically, performance increases with age to reach adult values in older children. To further assess age differences in SRT, each HINT condition was submitted to a one-way ANOVA. A significant effect of age was found in all conditions [quiet—F(5,74) ⫽ 37.710, p ⬍ 0.001; NF—F(5,74) ⫽ 18.531, p ⬍ 0.001; NS—F(5,74) ⫽ 106.126, p ⬍ 0.001], a finding consistent with the Nilsson et al. (1996) study. The spatial separation advantage measure was also submitted to a one-way ANOVA, for which a significant effect of age was found [F(5,74) ⫽ 25.883, p ⬍ 0.001]. Post hoc analyses were performed for each HINT condition and for the spatial separation advantage

Spatial separation advantage

Age groups being compared

P

6 and 8 6 and 10 6 and 12 6 and adults 8 and 12 8 and adults 10 and 12 10 and adults 6 and 8 6 and 10 6 and 12 6 and adults 8 and adults 10 and adults 6 and 8 6 and 10 6 and 12 6 and adults 8 and 10 8 and 12 8 and adults 10 and 12 10 and adults 6 and 10 6 and 12 6 and adults 8 and 12 8 and adults 10 and 12 10 and adults

0.005 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 0.006 0.001 0.014 ⬍0.001 ⬍0.001 ⬍0.001 0.001 0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 0.003 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 0.014 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 0.001 0.001

458

VAILLANCOURT

ET AL.

/ EAR & HEARING, VOL. 29, NO. 3, 453–466

TABLE 2. Regression coefficients for various types of trend lines Measure

Linear

Polynomial

Logarithmic

Quiet Noise front Noise side Spatial separation advantage

0.9696 0.9059 0.9984 0.9234

0.9717 0.9917 0.9988 0.9844

0.9513 0.9554 0.9826 0.8627

always statistically significantly different from one another, performance in the 12-yr olds was consistently significantly better than that of the 6-yr olds, justifying the need to establish a trend for performance as a function of age between these groups. Furthermore, no statistically significant difference was found between the performances of 12-yr olds and adults, eliminating the need to establish correction factors for ages beyond 12. Correction Factors • One objective of the study is to provide correction factors that can account for the effect of age on SRTs by comparing mean performances for the various age groups with adult performances. As this study included only a limited number of age groups, interpolation by means of regression equations must be used to predict performance for untested groups of children. Various regression and trend lines were used to fit the data between children aged 6 and 12, including linear, polynomial, and logarithmic, for which the regression coefficients are found in Table 2. The mean group scores were fit by means of a second-order polynomial (Figs. 2, 3) because it provided the best regression coefficients, a finding similar to that of Nilsson et al. (1996). The regression equations in Figures 2 and 3 were used to predict mean performance for all ages between 6 and 12. The correction factors, listed in Table 3, were obtained by subtracting from these predicted mean scores the measured performance

for adults. For comparative purposes, those obtained with the English version of the HINT for children (Nilsson, et al., 1996) are also included for the quiet, NF, and NS conditions. As can be seen, the correction factors are similar across both languages, indicating that the effect of age on speech reception abilities is similar for native English and French speakers. Indeed, apart from the quiet condition for 6- and 7-yr olds, all cross-language differences are less than 1 dB. A mean difference between the English and French correction factors across age groups of 0.9, 0.4, and 0.2 was found for the quiet, NF, and NS, respectively. The largest mean difference is noted for the quiet condition, where individual cross-language differences are greater for the 6-yr-old group than older groups. Normative Approach • Using the approach proposed by Nilsson et al. (1996), norms for a given soundfield are established by initially measuring adult performance in the soundfield. The measured mean adult SRTs are entered into the HINT software, which generates the adult norms and percentile rankings for the soundfield in question using the entered data in combination with SDs previously established with larger samples of adults with normal hearing. The correction factors are thereafter applied to these norms and percentile rankings to generate age-specific norms and rankings for children. Such an approach is based on the assumption that although mean scores for children are higher than that of adults, the dispersion of scores around the mean is similar in both groups. To verify this assumption and motivate the use of adult SDs already established with the Canadian French HINT, using larger samples, in generating age-specific norms for children, it must be demonstrated that (a) variance (␴2) among the scores does not change as a function of age, and (b) variance in previously established data is similar to variance obtained in this study.

TABLE 3. Age-specific correction factors for the English (Nilsson, et al., 1996) and French versions of the Hearing in Noise Test (HINT) for children Correction factors (dB) for ages Condition Quiet

Noise front

Noise side

Spatial separation advantage

Language

6

7

8

9

10

11

12

English French Difference across languages English French Difference across languages English French Difference across languages French

10.1 7.2 2.9 2.4 2.3 0.1 4.3 4.4 0.1 2.1

7.4 6.3 1.1 2.2 1.9 0.3 3.7 3.9 0.2 2.0

5.2 5.3 0.1 1.9 1.5 0.4 3.1 3.3 0.3 1.8

3.5 4.2 0.7 1.7 1.2 0.5 2.6 2.7 0.1 1.5

2.3 3.1 0.8 1.5 1.0 0.5 2.1 2.1 0.1 1.1

1.4 1.9 0.5 1.3 0.9 0.4 1.6 1.5 0.1 0.7

0.8 0.7 0.1 1.1 0.8 0.3 1.2 0.9 0.3 0.1

VAILLANCOURT

ET AL.

/ EAR & HEARING, VOL. 29, NO. 3, 453–466

Regarding the variance among scores as a function of age, the SDs obtained in this study did not follow a specific trend of increase or decrease as a function of age (as seen in Figs. 2, 3), and at first glance seemed relatively stable across age groups and comparable with adult values. Using the F statistic, a verification of the null hypothesis (␴12 ⫽ ␴12) was performed at ␣ ⫽ 0.05 in each of the HINT conditions by comparing the SDs for each group of children with the SD in the adult group (6-adults; 8-adults; 10-adults; 12-adults) and also by comparing the SDs obtained in each successive age groups (6 – 8, 8 –10, and 10 –12). With the exception of two comparisons (6 – 8 and 8 –10 for the quiet condition, where the SD was less for the 8-yr olds than the two other groups), the null hypothesis could not be rejected as Fobs (s12/s22) fell within the 95% confi(n1 ⫺ 1) dence interval delimited by F (n2 ⫺ 1, ␣ / 2) ⫽ 3.12 and (n1 ⫺ 1) F (n2 ⫺ 1, 1 ⫺ ␣ / 2) ⫽ 0.32. The assumption of equal variance was therefore verified in most comparisons, thereby justifying using adult SDs to yield percentile rankings in children. Indeed, no significant difference was found between variance in the 6-yrold children and variance in adult scores in any of the HINT conditions. A similar analysis was performed to compare the variance in scores in a group of adults previously tested to establish norms for the Canadian French HINT (Vaillancourt, et al., 2005) and the variance in scores obtained with the current sample of adults tested with the Canadian French HINT for children. Again, no significant difference in variance was found in any of the HINT conditions. These analyses suggest that one can estimate the SD in normalhearing children as being identical to that observed for performance in a larger number of normally hearing adults, as proposed by Nilsson et al. (1996). Previous work with the adult versions of the English and French HINT has established the population distributions and shows that SDs are similar for headphone and soundfield measures (⬃3.0 dB for the quiet condition and 1–1.5 dB for noise conditions) (Laroche, et al., 2005; Nilsson & Soli, 1994, Soli & Vermiglio, 1999; Vaillancourt, et al., 2005). More specifically, SDs for the French HINT are 3.8, 1.1, and 0.9 dB for the quiet, NF, and NS conditions, respectively, with a 1.2 dB SD for the spatial separation advantage. Greater variability in quiet can be attributed to differences in hearing sensitivity among normal-hearing subjects, differences that would only affect SRT measures in quiet (Nilsson, et al., 1994; Vaillancourt, et al., 2005). The already established SDs, together with the correction factors established during this study, can be applied to adult performances to generate agespecific norms for a given soundfield. Norms can

459

therefore be derived for any given soundfield test setting by measuring performance in a sample of normal-hearing adults for each HINT condition (to correct the headphone mean performance and percentile rankings for specific room effects) and then applying the age-specific correction factors found in Table 3.

Experiment 2—Validation of Normative Protocol With a Group of 9-Yr-Old Children Subjects • A total of 28 native French-speaking subjects with normal hearing who speak French on a daily basis took part in the validation study, including thirteen 9-yr olds and 15 adults between the ages of 18 and 30 (mean age ⫽ 22). Measured adult performance in this audiometric room served to correct the adult headphone norms for this particular soundfield before the application of correction factors. Nine-yr olds were chosen to represent an untested age group in the previous experiment. All subjects met the inclusion criteria described earlier and were required to read an information letter, sign a consent form, and fill out a hearing history questionnaire before testing. Procedure • Testing apparatus and procedures were similar to those used during the establishment of mean performances and correction factors, with the exception that testing was performed in a different soundfield (IAC audiometric room: 2.9 ⫻ 5.6 m2). After the hearing screening, three SRTs were obtained in each HINT condition using 10-sentence lists. In addition, the NF condition was repeated to assess test–retest reliability. As testing repetition requires three additional lists and the test material is composed of a total of 17 ten-sentence lists, test–retest measures were limited to only one HINT condition. Results Validity of Correction Factors • For each HINT condition, the best two of three SRTs were averaged to produce the subject’s overall score. NS and spatial separation advantage were also computed. Correction factors established in the previous experiment were then applied to adult performance measured in the current sample to predict mean SRTs for a group of 9-yr old children, which are compared with actual measured performance in Table 4. Apart from the quiet condition, all differences between predicted and measured mean SRT are less than 0.5 dB. Again, the large difference in quiet could potentially be attributed to the generally greater range in scores for this condition, to slight differences in pure-tone thresholds in the two samples (Soli & Nilsson, 1994; Vaillancourt, et al., 2005), or to differences in the ambient noise of the soundfield used to establish the

460

VAILLANCOURT

ET AL.

/ EAR & HEARING, VOL. 29, NO. 3, 453–466

TABLE 4. Mean performance for samples of native French-speaking normal hearing adults and 9-year old children and comparison of predicted and measured speech reception thresholds (SRT) for the group of children Mean measured SRT (dBA or dB S/N) Condition

Adults (N ⫽ 15)

9-Year old group (N ⫽ 13)

Mean predicted SRT for 9-year old children (dBA or dB SNR)

Difference between predicted and measured mean RTS (dB)

Quiet (dBA) Noise front (dB SNR) Noise side (dB SNR) Spatial separation advantage (dB)

18.9 ⫺2.7 ⫺10.4 7.8

18.5 ⫺1.7 ⫺7.7 6.0

23.1 ⫺1.5 ⫺7.7 6.3

4.6 0.2 0 0.3

correction factors and that used to validate those corrections. As individuals were screened for normal hearing at ⱕ15 dBHL and hearing thresholds were not measured per se, differences in pure-tone thresholds between samples or between age groups were not documented. Despite the underlying source of this difference, it implies that one cannot accurately predict mean performance in quiet using the correction factors. Reliability of SRT Measures • Using the testing protocol outlined in this study, the reliability of SRT measures can be expressed as the within-subjects SD of repeated measurements on two trials (test– retest reliability). The within-subjects SD of repeated measurements, ␴w, is calculated as follows:

␴␻ ⫽

冑

冘 冘 (x n

scores from list 1 to list 3 in a given condition was noted despite the fact that in many instances, the best two scores were obtained with the last two sentence lists. Indeed, a one-way ANOVA failed to reveal a significant effect of list position on scores. A similar analysis was performed on test and retest scores in the NF condition, for those subjects who were submitted to test–retest measures. Again, no systematic improvement in scores was noted from test to retest. It should however be noted that a particular list was never presented more than once to a given subject during these experimentations.

DISCUSSION Establishment of Age-Specific Norms

k

i, j

⫺ ␮ i) 2

i⫽1 j⫽1

n(k ⫺ 1)

where xi,j is the jth threshold of the ith subject, ␮i is the mean of the thresholds provided by the ith subject, k is the number of trials (k ⫽ 2 for test and retest), and n is the number of subjects. The reliability of SRT measures was computed using data gathered with subjects submitted to repeated measures of the NF condition, within the same testing session (9 yrs old and adult subjects from experiment 2). Because the null hypothesis (␴12 ⫽ ␴12) could not be rejected using the F statistic, variance among both groups was assumed to be equal and data were pooled to yield an overall within-subjects SD ␴w of 0.7 dB for the NF condition. Consequently, there is a probability of 0.95 that a SRT based on the average of the best two of three individual scores will fall within 1.96 ␴w of the “true” threshold. In addition, SRTs calculated in this manner that differ by more than 1.96 ␴w can be considered significantly different (p ⬍ 0.05). Learning Effect • During data analysis, the average performance across all subjects was also calculated for each list position (list 1, list 2, and list 3) in each condition. No systematic improvement in

The objective of this study was to provide a protocol for the establishment of age-specific normative data for the French HINT-C to facilitate its clinical use and allow comparison of an individual child’s performance with that of age-matched normal-hearing children. Percentiles for performance in each HINT condition can also be established to rank the performance of a child relative to that of an age-matched group. Using the current protocol, based on previous work by Nilsson et al. (1996), one avoids testing a large number of children of different ages, a generally laborious and time-consuming task. Instead, a limited number of normal hearing adults (⬃15–20) are tested in each HINT condition to correct the adult headphone norms and percentile rankings for the soundfield in question, and correction factors are then applied to generate age-specific norms and percentile rankings. Users are therefore only required to enter into the HINT software the mean adult performance values obtained in their soundfield and the system will automatically derive the adult norms, age-specific children norms, and percentile rankings for each population. Corrections for ages below 6 are inappropriate as the test material was developed based on vocabulary representative of 6-yr-old children, whereas corrections for ages beyond 12 are unnecessary. Indeed,

VAILLANCOURT

ET AL.

/ EAR & HEARING, VOL. 29, NO. 3, 453–466

the mean SRT in each HINT condition was found to improve with age until reaching adult performances at age 12. Consequently, the adult version of the French HINT could be used to assess speech reception abilities in children 12 yrs or older. Similarly, children performance on the English version reached adult-like values at age 13 (Nilsson, et al., 1996). The correction factors were obtained by comparing the mean performance of children on the HINT-C with that of adults using the same material, in the same soundfield. The approach is therefore somewhat different than that used by Nilsson et al. (1996), where children performance on the HINT-C was compared with adult performance on the adult version of the HINT, in the same soundfield. However, additional testing conducted with a group of 15 normally hearing French speakers demonstrated that adult performance on the HINT-C and adult HINT did not significantly differ [repeated-measures ANOVA: F(1,14) ⫽ 0.393, p ⫽ 0.541]. The correction factors established in the current study can therefore be applied to measured adult performance on the adult version of the HINT, thereby limiting the need to test two independent samples of normally hearing adults in a given soundfield; one group tested with the adult version to correct the adult headphone norms for that soundfield (establishment of adult normative data), and the other group tested with the children’s version to establish age-specific normative data. It was demonstrated in this study (experiment 2) that correction factors were effective in predicting mean SRTs for a group of previously untested children in all but the quiet condition. The difference between predicted and measured performances was less than 0.5 dB for the noise conditions but exceeded 4 dB for the quiet condition, a difference probably attributable in part to the generally greater variability of scores in this condition, differences in pure-tone thresholds, or differences in ambient noise between soundfields. Unfortunately, hearing thresholds were not measured in this study as screening was performed at 15 dBHL, and can therefore not be used to explain the underlying source of this difference. Mean performance in quiet can therefore not be predicted accurately using the correction factors. Not surprisingly, the correction factors established in this study closely mirror those obtained with English-speaking children (Nilsson, et al., 1996), indicating that maturational processes that could account for the improvement of speech reception abilities with age occur at a similar rate in both populations. Apart from the quiet condition for the

461

6- and 7-yr old children, cross-language differences in correction factors were less than 1 dB.

Effect of Age on Spatial Separation Advantage The effect of age on performance was not limited to SRT, but was also reflected in the spatial separation advantage. Such results would suggest that young children do not make as effective use of the advantage provided by the spatial separation of speech and noise as older children and adults, indicating that maturation of binaural processes involved in unmasking would continue to evolve over longer periods than other processes that do not tap into higher-order processes. Using a masking-level difference paradigm for pure tones, Hall et al. (2004) demonstrated improvements in binaural advantage with age in their sample (5 yr 1 mo to 10 yr 8 mo) and attributed their results to a developmental increase in binaural temporal resolution. With regards to the spatial separation advantage exhibited in the adult sample of the current study (mean spatial separation advantage ⫽ 6.8 dB), results concur with those of other investigations that have shown a spatial separation advantage ranging between 5 and 10 dB for a 90° separation between the signal and masker (e.g., Arbogast, 2002, 2005; Bronkhorst & Plomp, 1988; Cameron, et al., 2006a). Maturational effects on the ability to use spatial cues have however not been consistently reported in the literature. Despite a trend of decreasing SNR with age required to just understand a target story in the presence of distracter sentences spoken by the same voice or a different voice and simultaneously presented at either 0° or ⫾90° azimuth, Cameron et al. (2006b) failed to show a significant effect of age for spatial advantage, defined as the difference in thresholds between the 0° and ⫾90° conditions using a same voice paradigm, in their sample of 7-, 8-, and 9-yr olds and adults. In contrast, there was a significant effect of age on the tonal advantage (difference in thresholds between the same voice and different voice for the 0° condition) and total advantage (difference in thresholds between the same voice 0° and different voice ⫾90°). On the other hand, as shown by Johnstone and Litovsky (2006), the amount of spatial release from masking (SRM) is somewhat dependent on the type of interferer used, especially in children. Compared with adults, children aged 5- to 7-yr had higher SRT across all conditions studied, significantly larger SRM in a reversed-speech distracter and significantly smaller SRM in a modulated-noise distracter. Such dependence of SRM on the type of distracter could in part explain differences across studies.

462

VAILLANCOURT

ET AL.

/ EAR & HEARING, VOL. 29, NO. 3, 453–466

Effect of Age on Speech Recognition Performance Sample size in the current study may not have been large enough to establish statistically significant differences in mean SRTs across the various age groups, especially in the NF condition. However, significant differences between the youngest and oldest age groups supported establishing a trend to study the effect of age on SRTs, which increased with age until reaching adult values in children aged 12 yr. A lack of experience with language could account for poorer speech recognition performances in young children. Having less language experience, it can be hypothesized that young children may not be able to use contextual (semantic and syntactic) cues as effectively as older children and adults to recognize sentences, a difference that would disappear with increased experience (typically with age and frequency of use). Young children could therefore be likened to individuals listening to speech in their second language (L2), for which they have limited experience. van Wijngaarden (2003) has investigated the effects of L2 on speech recognition and demonstrated that not only did individuals listening to L2 typically require a higher SNR (in average 1–7 dB) than their counterparts listening in their native tongue (L1) for equal sentence recognition performance, but the slope of their psychometric function relating percent performance and SNR was also significantly shallower, these effects being greatest in those with the most limited L2 experience. Limited experience with English in childhood was also demonstrated by Nilsson et al. (1996) to increase SRTs in all HINT conditions. Using the English HINT, Lamothe et al. (2002) nonetheless found no significant difference in SRTs between native English speakers and bilingual French speakers who have acquired English before the age of 11. Indeed, individuals who acquired L2 at an early age typically have better speech recognition performances than those who acquired it later in life (Bahrick, et al., 1994; Mayo, et al., 1997; Meador et al., 2000). Moreover, a study by Florentine (1985), in which the difference between L2 and L1 listeners was shown to be significantly greater for high predictability sentences than low predictability sentences supports findings that L2 listeners are unable to make as effective use of contextual cues as L1 listeners (Florentine, et al., 1984; Lecumberri, et al., 2006; Mayo, et al., 1997; Shimizu, et al., 2002; van Wijngaarden, 2003). It would therefore seem that individuals with limited language experience (i.e., L2 listeners or younger children) would have lower SRTs and would not benefit from increases in SNR to the same extent

as those who master the language (i.e., adult L1 listeners or fully bilingual individuals). Although consensus has not been reached with regards to the underlying processes, improvements in speech recognition performance with age have been reported repeatedly in the literature, using various speech stimuli (monosyllabic words, spondees, words, sentences, stories) (e.g., Blandy & Lutman, 2005; Cameron, et al., 2006b; Eisenberg, et al., 2000, 2002; Hnath-Chisolm, et al., 1998; Johnson, 2000; Lebel & Picard, 1997; Picard & Bradley, 2001; Stelmachowicz, et al., 2000; Stuart, 2005). Studies comparing the performance of children across various age groups have demonstrated that younger children (5 and 6 yrs old) have greater difficulty in recognizing words (monosyllables, spondees, trochee, and trisyllabic words) presented at a given SNR than older children (Jamieson, et al., 2004) and require on average a 7 dB increase in SNR to reach performance levels similar to those obtained by older children during monosyllabic and disyllabic word identification tasks (Bradley & Sato, 2004; Picard & Bradley, 2001). Using the Bamford-Kowal-Bench sentences (Bench, et al., 1979) presented dichotically under headphones in a speech-shaped noise, Blandy and Lutman (2005) found that despite having hearing threshold levels better than or equal to those of young adults, 7-yr olds generally have a significantly poorer ability to recognize speech in noise. A 3 dB difference in SNR was found between the mean performances of young adults and 7-yr-old children to recognize 70% of the sentences presented. The authors concluded that the ability to recognize speech in noise is not fully developed in 7-yr olds, most likely attributable to immature central auditory processing. Despite differences in methodology (i.e., scoring procedure, definition of SRT) and material used, this finding is somewhat similar to the 2 dB correction factor established for the NF condition of both the French and English versions of the HINT for 7-yr old children. Findings of the current study are also in good agreement with the literature on age effects in auditory processing abilities. Performance on a variety of auditory tasks has been demonstrated to increase with age to reach adult-like values in adolescence or past 10 yrs (Eisenberg, et al., 2000; Elliott, 1979; Hartley, et al., 2000; Johnson, 2000; Neuman & Hochberg, 1983; Stollman, et al., 2004; Stuart, 2005), a finding consistent with results obtained during SRT measurements using both the French and English versions of the HINT for children, where performance reached adult values at age 12 and 13, respectively.

VAILLANCOURT

ET AL.

/ EAR & HEARING, VOL. 29, NO. 3, 453–466

It generally seems that the type of auditory task highly influences the age at which children reach adult-like performance on tests of speech perception (Hnath-Chisolm et al., 1998; Stuart, 2005). The ability to recognize speech in noise is a complex auditory task entailing not only the identification and discrimination of speech sounds and peripheral sensitivity to speech, but also higher-level auditory processing to separate the speech signal from the noise. It is therefore not surprising that maturational effects would come into play for children at least until adolescence, as previous findings have shown that some age-related changes in cortical auditory function and auditory processing abilities extends into adolescence in normally developing individuals (Cunningham, et al., 2000; Johnstone, et al., 1996; Keith, 1995; Neijenhuis, et al., 2002; Ponton et al., 1996, 2000). Overall, behavioral maturation in speech reception abilities appears to parallel the developmental time course of the structural and electrophysiological development of the auditory cortex, reaching maturity in adolescence (Moore & Gaung, 2001; Moore, 2002; Eggermont & Ponton, 2003). A recent longitudinal study following a group of 6-yr old children showed an improvement with increasing age (up to 12–13 yr) in all auditory processing tasks surveyed (filtered speech task, binaural fusion task, frequency pattern task, and duration pattern task), with the exception of the speech recognition in noise task (Stollman, et al., 2004). The lack of improvement noted for speech recognition could however potentially be attributable to the simplicity of the task and material used (lists of CVC monosyllables) and/or ceiling effects associated with percent correct scoring. A longitudinal study following a group of 6-yr olds into adulthood using a variety of speech stimuli (monosyllables, spondees, words, and sentences) and tasks varying in degree of difficulty, in addition to electrophysiological measures, would be warranted to further shed light on maturational effects on the ability of children to recognize speech in noise. Other Future Work The within-subject SD of repeated measures was calculated from subject data gathered during validation of the correction factors (experiment 2). Despite no significant difference in within-subject SD between the two groups submitted to repeated measures (9-yr olds and adult sample), it should further be verified that test–retest variability is not age-dependent. No learning effect was found in the current study when analyzing test–retest scores obtained within the same testing session, using different lists. No investigation into the presence of a memory effect

463

was performed using repeated presentations of the same test material and longer test–retest intervals, a phenomenon that should be assessed in further work. Should a learning effect exist in such cases, more lists would need to be generated and/or the optimal test–retest interval would need to be determined, ensuring that children’s performances would not improve at retest solely as a result of remembering the sentences. Comparing performance with high and low predictability material in children of various ages could help substantiate the role of language experience in accounting for increases in speech recognition performance as a function of age into adolescence. Further work investigating the effect of age on the slope of the psychometric function in children is also warranted. In line with studies on speech recognition in a second language, it can be hypothesized that children having more limited language experience (younger children) would have shallower slopes and hence would not benefit from increases in SNR as much as older children and adults. Such findings could have considerable implications in classroom or FM system applications. Finally, to determine whether or not the effects of L2 are similar in adults and children, studies investigating the intelligibility of non-native speech in children should be carried out.

CONCLUSION A single set of age-specific correction factors to be added to adult mean performance in a given soundfield have been determined in this study to facilitate the process of establishing normative data for children of various ages in each clinical test site. These correction factors were also shown to be effective in predicting performance in a group of children in all HINT conditions apart from the quiet condition. To obtain age-specific norms and percentile rankings, users must measure performance in all HINT conditions in a sample of 15 to 20 normal-hearing adults, using the adult HINT material (one 20sentence list per condition), and enter measured mean SRTs in the HINT software, thereby generating soundfield norms and percentile rankings for adults. The software then applies the correction factors to these values to generate age-specific norms and percentile rankings for the same soundfield. Using the age-correction factors from this study to establish age-specific norms for normal-hearing children, a hearing-impaired child’s SRTs on HINT could be compared with that of age-matched normally hearing children to rank the child’s performance and quantify impairment relative to SNR

464

VAILLANCOURT

ET AL.

/ EAR & HEARING, VOL. 29, NO. 3, 453–466

loss (amount of additional SNR required to achieve the same performance). For example, a child whose scores on HINT differ from age-specific normative data by 4 dB has a 4 dB SNR loss and thereby requires a SNR of 4 dB higher than that required by normal-hearing counterparts to reach the same level of performance. Such knowledge could prove helpful in determining adequate rehabilitation strategies (i.e., hearing aids, FM systems, cochlear implantation, communication strategies, preferential seating in classrooms), especially with regards to improving the SNR. Where normal language development is however in doubt, the child’s performance could not only be compared with that of age-matched children, but also to performance in an age group with similar language experience, such as determined by a language assessment (i.e., language inventory). Finally, when testing hearing-impaired individuals (adults and children), the noise level can be increased beyond 65 dBA when the SRT in quiet exceeds 45 dBA, thereby ensuring that audibility issues do not hinder speech recognition in noise abilities.

ACKNOWLEDGMENTS Isabelle Carrie`re assisted in the data acquisition. The authors acknowledge the schools that assisted in the recruitment process and the children and their parents who participated in this study. Such precious collaboration allowed establishing a normative protocol that will greatly facilitate clinical practice. This research was supported by a grant from the Consortium National de Formation en Sante´ (CNFS). Address for correspondence: Ve´ronique Vaillancourt, Faculty of Health Sciences, Room 1117, Audiology and Speech-Language Pathology Program, University of Ottawa, 451 Smyth Road, Ottawa, Ontario, Canada K1H 8M5. E-mail: [email protected]. uottawa.ca. Received March 1, 2007; accepted November 20, 2007.

REFERENCES American National Standards Institute. (1996). American national standard: specification for audiometers (ANSI S3.61996). New York: Acoustical Society of America. Arbogast, T. L., Mason, C. R., & Kidd, G. Jr. (2002). The effect of spatial separation on informational and energetic masking of speech. J Acoust Soc Am, 112, 2086 –2098. Arbogast, T. L., Mason, C. R., & Kidd, G. Jr. (2005). The effect of spatial separation on informational masking of speech in normal-hearing and hearing-impaired listeners. J Acoust Soc Am, 117, 2169 –2180. Bahrick, H. P., Hall, L. K., Goggin, J. P., et al. (1994). Fifty years of language maintenance and language dominance in bilingual Hispanic immigrants. J Exp Psychol Gen, 123, 264 –283. Bellis, T. J. (2003). Assessment and management of central auditory processing disorders in the educational setting: from science to practice. Clifton Park, NY: Thomson/Delmar Learning. Bench, J., Kowal, A., & Bamford, J. (1979). The BKB (BamfordKowal-Bench) sentence lists for partially-hearing children. Br J Audiol, 13, 108 –112.

Blandy, S., & Lutman, M. (2005). Hearing threshold levels and speech recognition in noise in 7-year-olds. Int J Audiol, 44, 435– 443. Bosman, A. J. (1989). Speech perception by the hearing impaired. Doctoral dissertation, University of Utrecht, The Netherlands. Bradley, J. S., & Sato, H. (2004). Speech recognition by grades 1, 3 and 6 children in classrooms. Can Acoust, 32, 26 –27. Bronkhorst, A. W., & Plomp, R. (1988). The effect of head-induced interaural time and level differences on speech intelligibility in noise. J Acoust Soc Am, 83, 1508 –1516. Bronkhorst, A. W., & Plomp, R. (1990). A clinical test for the assessment of binaural speech perception in noise. Audiology, 29, 275–285. Bronkhorst, A. W., & Plomp, R. (1992). Effect of multiple speechlike maskers on binaural speech recognition in normal and impaired hearing. J Acoust Soc Am, 92, 3132–3139. Cameron, S., Dillon, H., & Newall, P. (2006a). Development and evaluation of the listening in spatialized noise test. Ear Hear, 27, 30 – 42. Cameron, S., Dillon, H., & Newall, P. (2006b). The Listening in Spatialized Noise test: normative data for children. Int J Audiol, 45, 99 –108. Chermak, G. D., & Musiek, F. E. (1997). Central auditory processing disorders: new perspectives. San Diego, CA: Singular Publishing Group. Cox, R. M., Alexander, G. C., & Gilmore, C. (1987). Development of the connected speech test (CST). Ear Hear, 8, 119 –126. Crandell, C. (1991). Individual differences in speech recognition ability: implications for hearing aid selection. Ear Hear, 12, 100S–108S. Cunningham, J., Nicol, T., Zecker, S., et al. (2000). Speech-evoked neurophysiologic responses in children with learning problems: development and behavioural correlates of perception. Ear Hear, 21, 554 –568. Dubno, J. R., Dirks, D. D., & Morgan, D. E. (1984). Effects of age and mild hearing loss on speech recognition in noise. J Acoust Soc Am, 76, 87–96. Duncan, R. C., Knapp, R. G., & Miller, M. C. (1983). Introductory biostatistics for the health sciences. (2nd ed.). New York, NY: Willey. Duquesnoy, A. J. (1983). Effects of a single interfering noise or speech source upon the binaural sentence intelligibility of aged persons. J Acoust Soc Am, 74, 739 –743. Eggermont, J. J., & Ponton, C. W. (2003). Auditory-evoked potential studies of cortical maturation in normal hearing and implanted children: correlations with changes in structure and speech perception. Acta Otolaryngol, 123, 249 –252. Eisenberg, L. S., Martinez, A. S., Holowecky, S. R., et al. (2002). Recognition of lexically controlled words and sentences by children with normal hearing and children with cochlear implants. Ear Hear, 23, 450 – 462. Eisenberg, L. S., Shannon, R. V., Martinez, A. S., et al. (2000). Speech recognition with reduced spectral cues as a function of age. J Acoust Soc Am, 107, 2704 –2710. Elliott, L. L. (1979). Performance of children aged 9 to 17 years on a test of speech intelligibility in noise using sentence material with controlled word predictability. J Acoust Soc Am, 66, 651– 653. Fels, J., & Vorla¨nder, M. (2004). Artificial heads for children. Proceedings of the 18th International Congress on Acoustics, Japan, 3457–3458. Florentine, M. (1985). Non-native listeners’ perception of American-English in noise. Proceedings of Internoise, 85, 1021–1024. Florentine, M., Buus, S., Scharf, B., & Canevet, G. (1984). Speech reception thresholds in noise for native and non-native listeners. J of Acoust Soc Am, 75, 84 – 84.

VAILLANCOURT

ET AL.

/ EAR & HEARING, VOL. 29, NO. 3, 453–466

Gatehouse, S., & Noble, W. (2004). The speech, spatial and qualities of hearing scale (SSQ). Int J Audiol, 43, 85–99. Gelfand, S. A., Ross, L., & Miller, S. (1988). Sentence reception in noise from one versus two sources: effects of aging and hearing loss. J Acoust Soc Am, 83, 248 –256. Hagerman, B. (1982). Sentences for testing speech intelligibility in noise. Scand Audiol, 11, 79 – 87. Hall, J. W., Buss, E., Grose, J. H., et al. (2004). Developmental effects in the masking-level difference. J Speech Lang Hear Res, 47, 13–20. Hall, J. W., & Mueller, G. H. (1997). Audiologist Desk Reference Volume 1. Diagnostic audiology principles, procedures and practices. San Diego, CA: Singular Publishing Group Inc. Hartley, D. E., Wright, B. A., Hogan, S. C., et al. (2000). Agerelated improvements in auditory backward and simultaneous masking in 6- to 10-year-old children. J Speech Lang Hear Res, 43, 1402–1415. Hnath-Chisolm, T. E., Laipply, E., & Boothroyd, A. (1998). Age-related changes on a children’s test of sensory-level speech perception capacity. J Speech Lang Hear Res, 41, 94 –106. Jamieson, D. G., Kranjc, G., Yu, K., et al. (2004). Speech intelligibility of young school-aged children in the presence of reallife classroom noise. J Am Acad Audiol, 15, 508 –517. Johnson, C. E. (2000). Children’s phoneme identification in reverberation and noise. J Speech Lang Hear Res, 43, 144 –157. Johnstone, S. J., Barry, R. J., Anderson, J. W., et al. (1996). Age-related changes in child and adolescent event-related potential component morphology, amplitude and latency to standard and target stimuli in an auditory oddball task. Int J Psychophysiol, 24, 223–238. Johnstone, P. M., & Litovsky, R. Y. (2006). Effect of masker type and age on speech intelligibility and spatial release from masking in children and adults. J Acoust Soc Am, 120, 2177– 2189. Kalikow, D. N., Stevens, K. N., & Elliot, L. L. (1977). Development of a test of speech intelligibility in noise using sentence materials with controlled word predictability. J Acoust Soc Am, 61, 1337–1351. Keith, R. W. (1995). Development and standardization of SCAN-A: test of auditory processing disorders in adolescents and adults. J Am Acad Audiol, 6, 286 –292. Killion, M. C., & Niquette, P. A. (2000). What can the pure-tone audiogram tell us about a patient’s SNR loss? Hear J, 53, 46 –53. Killion, M. C., & Villchur, E. (1993). Kessler was right partly: but SIN test shows some aids improve hearing in noise. Hear J, 46, 31–35. Kochkin, S. (2000). MarkeTrak V: “Why my hearing aids are in the drawer”: the consumers’ perspective. Hear J, 53, 34 – 41. Kochkin, S. (2002). 10-Year customer satisfaction trends in the US hearing instrument market. Hear Rev, 9, 17–21. Lamothe, J., Gascon, C., Larivie`re, M., et al. (2002). Normalisation du hearing in noise test (HINT) aupre`s d’une population francophone bilingue et d’une population anglophone. Revue d’orthophonie et d’audiologie, 26, 81– 89. Laroche, C., Gigue`re, C., Vaillancourt, V., et al. Development and validation of hearing standards for Canadian Coast Guard seagoing personnel and Conservation and Protection seagoing and land-based personnel, Phase II. Final Report, Submitted to Canadian Coast Guard under Contract No. F7053– 000009, (March 31st, 2005), 64 p. (⫹14 Appendices). Laroche, C., Soli, S. D., Gigue`re, C., et al. (2003). An approach to the development of hearing standards for hearing-critical jobs. Noise Health, 6, 17–37.

465

Laroche, C., Vaillancourt, V., Melanson, C., et al. (2006). Adaptation du HINT (Hearing in Noise Test) pour les enfants francophones canadiens et donne´es pre´liminaires sur l’effet d’aˆge. Revue d’Orthophonie et d’Audiologie, 30, 95–109. Lebel, C., & Picard, M. (1997). Influence du mode de re´ponse sur le seuil de reconnaissance de la parole chez l’enfant franc¸ais que´be´cois d’aˆge scolaire. Revue d’orthophonie et d’audiologie, 21, 17–27. Lecumberri, M. L. G., & Cooke, M. (2006). Effect of masker type on native and non-native consonant perception in noise. J of Acoust Soc Am, 119, 2445–2454. Leduc, R. (1997). Pour la re´ussite du de´pistage pre´coce et continue. Vanier, ON: Centre franco-ontarien de ressources pe´dagogiques. MacLoed, A., & Summerfield, Q. (1987). Quantifying the contribution of vision to speech perception in noise. Br J Audiol, 21, 131–141. MacLoed, A., & Summerfield, Q. (1990). A procedure for measuring auditory and audio-visual speech-reception thresholds for sentences in noise: rationale, evaluation, and recommendations for use. Br J Audiol, 24, 29 – 43. Mayo, L. H., Florentine, M., & Buus, S. (1997). Age of secondlanguage acquisition and perception of speech in noise. J Speech Lang Hear Res, 40, 686 – 693. Meador, D., Flege, J. F., & Mackay, I. R. A. (2000). Factors affecting the recognition of words in a second language. Bilingualism: Language and Cognition, 3, 55– 67. Medwestsky, L., Sanderson, D., & Young, D. (1999). A national survey of audiology clinical practices. Part 2: Evidence suggests need for “best practices” standards. Hear Rev, 6, 14 –22. Moore, J. K. (2002). Maturation of human auditory cortex: implications for speech perception. Ann Otol Rhinol Laryngol, 111, 7–10. Moore, J. K., & Guang, Y.-L. (2001). Cytoarchitectural and axonal maturation in human auditory cortex. J Assoc Res Otolaryngol, 2, 297–311. Mueller, G. H. (2001). Speech audiometry and hearing aid fittings: going steady or casual acquaintances?. Hear J, 54, 19 –29. Mueller, G. H. (2003). Fitting test protocols are “more honored in the breach than the observance”. Hear J, 56, 19 –26. Nabelek, A., & Pickett, J. (1974). Reception of consonants in a classroom as affected by monaural and binaural listening, noise, reverberation, and hearing aids. J Acoust Soc Am, 56, 628 – 639. Neijenhuis, K., Snik, A., Priester, G., et al. (2002). Age effects and normative data on a Dutch test battery for auditory processing disorders. Int J Audiol, 41, 334 –346. Neuman, A. C., & Hochberg, I. (1983). Children’s perception of speech in reverberation. J Acoust Soc Am, 73, 2145–2149. Nilsson, M., Soli, S. D., & Sullivan, J. A. (1994). Development of the hearing in noise test for the measurement of speech reception thresholds in quiet and in noise. J Acoust Soc Am, 95, 1085–1099. Nilsson, M. J., Soli, S. D., & Gelnett, D. (1996). Development of the hearing in noise test for children (HINT-C). House Ear Institute, April, 1–9. Picard, M. (2003). L’e´cole que´be´coise est-elle accessible a` tous les e´le`ves qui pre´sentent des difficulte´s d’e´coute? Fre´quences, 16, 7–11. Picard, M., & Bradley, J. S. (2001). Revisiting speech interference in classrooms. Audiology, 40, 221–244. Plomp, R. (1977). Acoustical aspects of cocktail parties. Acustica, 38, 186 –191. Plomp, R., & Mimpen, A. M. (1979). Improving the reliability of testing the speech reception threshold for sentences. Audiology, 18, 43–52.

466

VAILLANCOURT

ET AL.

/ EAR & HEARING, VOL. 29, NO. 3, 453–466

Ponton, C. W., Don, M., Eggermont, J. J., et al. (1996). Maturation of human auditory cortical function: differences between normal-hearing children and children with cochlear implants. Ear Hear, 17, 430 – 437. Ponton, C. W., Eggermnont, J. J., Kwong, B., et al. (2000). Maturation of human central auditory system activity: evidence from multi-channel evoked potentials. Clin Neurophysiol, 111, 220 –236. Rowland, J., Dirks, D., Dubno, J., et al. (1985). Comparison of speech recognition-in-noise and subjective communication assessment. Ear Hear, 6, 291–296. Ruscetta, M. N., Arjmand, E. M., & Sheila, R. P. (2005). Speech recognition abilities in noise for children with severe-to-profound unilateral hearing impairment. Int J Pediatr Otorhinolaryngol, 69, 771–779. Shimizu, T., Makishima, K., Yoshida, M., & Yamagishi, H. (2002). Effect of background noise on perception of English speech for Japanese listeners. Auris Nasus Larynx, 29, 121–125. Smoorenburg, G. F. (1992). Speech reception in quiet and in noisy conditions by individuals with noise-induced hearing loss in relation to their tone audiogram. J Acoust Soc Am, 91, 421– 437. Soli, S. D., & Nilsson, M. (1994). Assessment of communication handicap with the HINT. Hear Instrum, 45, 12–16.

Soli, S. D., & Vermiglio, A. (November 1999). Assessment of functional hearing ability for hearing critical jobs in law enforcement. A report prepared for the California Peace Officers Standards and Training Commission. Human Communication Sciences and Devices Department, House Ear Institute. Stelmachowicz, P. G., Hoover, B. M., Lewis, D. E., et al. (2000). The relation between stimulus context, speech audibility, and perception for normal-hearing and hearing-impaired children. J Speech Lang Hear Res, 43, 902–914. Stollman, M. H. P., van Velzen, E. C. W., Simkens, H. M. F., et al. (2004). Development of auditory processing in 6 –12year-old children: a longitudinal study. Int J Audiol, 43, 34 – 44. Stuart, A. (2005). Development of auditory temporal resolution in school-age children revealed by word recognition in continuous and interrupted noise. Ear Hear, 26, 78 – 88. Strom, K. (2003). The HR 2003 dispenser survey. Hear Rev, 10, 22–38. Vaillancourt, V., Laroche, C., Mayer, C., et al. (2005). Adaptation of the HINT (Hearing in Noise Test) for adult Canadian Francophone populations. Int J Audiol, 44, 358 –369. van Wijngaarden, S. J. (2003). The intelligibility of non-native speech. Doctoral thesis, Free University, Amsterdam.