Abstract
To investigate the effectiveness of computer-based auditory training on speech-in-noise perception in adults. With no language restriction, 11 databases were searched from 1990 to 2020. We included any clinical trial studies with concurrent comparison groups that examined the effectiveness of computer-based auditory training programs in adults. The primary outcome was a speech in noise perception that was estimated using the “difference pretest–posttest-control” index (dppc2). The risk of bias was assessed using the Cochrane collaboration tool for assessing the risk of bias in randomized trials. The certainty of the evidence was investigated using the GRADE in two primary outcomes. Twenty three studies were included in two subgroups based on primary outcome: 12 studies with speech perception threshold and 11 studies with speech-in-noise test scores. Computer-based auditory training resulted in a speech in noise perception improvement (dppc2: −0.69, 95%CI: −1.11 to −0.26; I2 = 69.6%, p = 0.00) and (dppc2: 0.71, 95%CI: 0.38–1.03, I2: 17.8%, p = 0.27) respectively in both subgroups. 19 studies were judged to have a high risk of bias and 3 studies had a low risk of bias and the strength of the evidence was low in both primary outcomes. This finding indicates that computer-based auditory training can be a moderately effective intervention for speech-in-noise perception in adults. However, due to the low quality of primary studies and the low certainty of the evidence, the results are not yet definite. Prospero registration number: CRD42021233193.
Supplementary Information
The online version contains supplementary material available at 10.1007/s12070-023-03920-0.
Keywords: Computer-based auditory training, Meta-analysis, Review, Speech-in-noise
Introduction
As reported by the World Health Organization (WHO), in 2020, 466 million people (6.1% of the global population) experienced hearing loss. This estimate is expected to reach 630 million by 2030[1]. The effect of hearing loss on personal life depends on the degree of hearing loss, lifestyle, communication needs, and other factors [2]. Hearing loss can result in isolation, depression, and many other cognitive problems [3]. However, a common complaint in individuals with hearing loss is speech perception in environments with multiple sources of sound [4]. Speech perception in noisy environments may be difficult even for people with normal hearing. 5–15% of those visiting audiology clinics, despite normal hearing threshold (≤ 25 dB HL, at 500 to 4000 Hz), reported speech in noise perception difficulties [5].
Hearing aids are the most common solution for hearing loss [6]. It seems that the speech-in-noise perception problem is a major factor involved in dissatisfaction with hearing aids [3]. Hearing aids amplify environmental sounds, but other interventions such as auditory training are still needed for listening and speech perception [6].
Since 1980, in addition to the prescription of hearing aids, audiologists have provided auditory rehabilitation services, including face-to-face auditory training [7]. Difficulty making an appointment with the audiologist, and financial and other issues led to the emergence of computer-based auditory training [8, 9]. There are different auditory training programs, including listening and communication enhancement (LACE), speech perception training system, ReedmyQuips (RMQ), and computer-based auditory phoneme discrimination training. The flexibility of these programs enables people to receive their training at home under comfortable conditions [10]. The opportunity for daily training and active participation of individuals in computer-based auditory training can improve the effectiveness of these programs [8]. Despite the different methods used in computer-based auditory training, it is essential to make the required adaptation to the level of the participant’s abilities and provide feedback to maintain interest and motivation [11]. Furthermore, the structure of these programs allows for close monitoring of the participant’s progress [10].
Previous systematic reviews showed computerized and non-computerized auditory training programs were effective in the speech perception and communication abilities of hearing-impaired individuals. However, due to the considerable methodological heterogeneity and the low quality of studies included in these systematic reviews, there is a strong need for conclusive evidence in the future body of research [6, 7].
As the existing literature showed, no conclusive systematic review has yet explored the effect of computer-based auditory training on adult speech-in-noise perception. Thus, the present study filled the gap and explored this issue using a more comprehensive approach to a meta-analysis of more databases. The primary outcome was to investigate the effect of computer-based auditory training on speech-in-noise perception in adults with and without hearing loss. The secondary outcomes were to explore the effects of computer-based auditory training on the quality of life and hearing disability, to test the potential heterogeneity of these studies, and find the potential causes.
Methods
The protocol of the present study was registered in the international prospective register of systematic reviews (PROSPERO) with identification number CRD42021233193. There was a protocol amendment: we excluded Single-group studies from the review: during the search, after ensuring a sufficient number of primary studies, we eliminated these types of studies because of their methodological weaknesses. Instead, studies with at least two groups (comparison and intervention) were included. The current study adhered to the 2020 Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement (online Appendix 1).
The primary search was done in PubMed using the two main components of the research question, the intervention (auditory training) and the outcome (speech-in-noise perception). To retrieve all synonyms, in addition to MeSH, we used expert opinions and previous primary and secondary studies. We searched the PubMed/MEDLINE, EMBASE, Web of Science, Google Scholar, Scopus, Cochrane Library, Cochrane Central Register of Controlled Trials (CENTRAL), Clinicaltrial.gov, International Standard Randomized Controlled Trial Number registry (ISRCT), International Clinical Trials Registry Platform (ICTRP), ProQuest for relevant theses/dissertations, Scopus and Web of Science for conference papers from 1990 to November 30, 2020. Reference lists of the included studies and the last six months’ content of the Journal of the Acoustical Society of America and Hearing Research were also searched. No limitation was set on the language of studies. We translated non-English papers [12, 13] using Google Translate. Online Appendix 2 shows the search syntax across all databases and the time of the search. The results were saved in Mendeley 1.19.4, and duplicates were removed. The title and abstract of papers were screened based on the inclusion and exclusion criteria. Thereafter, based on the full texts, the final papers were selected by two researchers (T.F. and R.T.) independently. The researchers held discussions to resolve disagreements; if required, a third expert was consulted to make the final decision. Cohen’s kappa coefficient (κ) was used to rate the inter-rater agreement [14]. All these phases were reported in the PRISMA flow diagram.
Inclusion Criteria
The inclusion criteria were based on PICOS: participants, intervention, comparison group, outcome, study type. These are described below:
Participants: adults (above 18 years old), both sexes included, with normal hearing or hearing loss, adults using or not using hearing aids or cochlear implants.
Intervention: auditory training not presented by an audiologist, performed on a PC, laptop, tablet, or telephone at home or in a clinic.
Comparison group: matched control not receiving any auditory training intervention.
Outcome: the primary outcome was described as an improvement in the speech-in-noise perception, represented in the two groups of speech perception threshold and speech-in-noise test score. The secondary outcomes included exploring the effect of computer-based auditory training on quality of life and hearing disability, assessing potential heterogeneity in primary studies, and finding its potential causes.
Study type: any clinical trial (pre- and post-design, two- or multiple-arm, parallel-group with or without randomization, repeated-measures, or cohort) was accepted. If the auditory training was examined alongside another intervention or if the results were reported in a separate group, the study was also accepted.
Data Extraction
A data extraction form was designed in Excel and completed by two researchers (T.F. and R.T.) independently. To resolve disagreements, discussions were held between the two researchers, and, if required, a third expert was consulted to make the final decision. The data extraction form included the first author’s name, publication year, type of study, sample size, participants’ demographic information (age, sex, hearing threshold, hearing aid, etc.), details of the auditory training program (stimuli, auditory task, processing type, duration), outcomes (speech-in-noise test, as well as questionnaires on quality of life and hearing disability). When the full text was not available, the researchers contacted the corresponding authors (three emails at 7–14-day intervals); if no response was received, the paper was excluded.
Measurement of Intervention Effect
The primary outcome was the assessment of speech-in-noise perception, reported as the speech perception threshold or speech-in-noise test score. The measurement unit for the speech perception threshold was decibels (dB HL) ranging from 0 to 100. The speech-in-noise test score was defined as the percentage of correct responses, ranging from 0 to 100. In studies with different time points to measure the speech-in-noise test, the first time point before and after the intervention was used as the pre-and post-intervention points, respectively [10, 15–17]. In primary studies in which more than two groups were presented, our strategy was to convert them into 2 × 2 groups (pairwise comparison: we considered one group as the reference and compared the other groups to it) for the meta-analysis [10, 17–21]. Furthermore, in a number of studies, there was more than one speech-in-noise test; for these, only one test was chosen [9, 15, 19, 22–24]. Therefore, the mean and standard deviation were extracted for both outcome measures. In a few studies [11, 19–21, 23, 25–27], these data were obtained from figures through a web-based tool (web plot digitizer) [28]. The effect size was measured using the “difference pretest–posttest-control” index (dppc21) [29, 30]. To calculate this index, a correlation coefficient of 0.3 was considered. The Campbell collaboration website was used for this aim [31]. A meta-analysis was done in Stata (version 14.2; Stata Corp, College Station, TX, USA). Due to the methodological heterogeneity among primary studies, we use the random effect model which is the more conservative combination model. The effect size of t intervention was interpreted as trivial (SMD2 < 0.19), weak (0.20 ≤ SMD ≤ 0.49), moderate (0.50 ≤ SMD ≤ 0.79), and high (SMD ≥ 0.80) [32]. In each primary outcome, a subgroup analysis was also carried out based on bottom-up or top-down approaches (the auditory training that included only low-level sensory processing is in the bottom-up category, and training that included both low-level sensory and high-level cognitive processing is in both processing categories), the presence of speech-in-noise test in the training, duration of training (week), the intensity of training (number of days per week), the place of training (lab or home), age range (< 30, 31–64, > 65), hearing level (normal, mild hearing loss, moderate and above hearing loss) according to the following criteria: normal hearing threshold ≤ 25 dB HL, mild hearing impairment = 26–40 dB HL, and moderate hearing impairment = 41–55 dB HL) [33]; use/nonuse of hearing aid/cochlear implant; study design (CT/RCT), quality score (low, medium, high), randomization, blinding of participants and researchers, and baseline similarity to important factors.
The secondary outcome measures were personal reports of the quality of life, and hearing disability, reported as the scores on the corresponding questionnaires. The included studies used different instruments to assess these outcomes. Therefore, the desired outcome criteria were different. Since the minimum conditions (absence of highly severe heterogeneity) were not feasible [14], a meta-analysis could not be conducted for secondary outcomes.
Assessment of Statistical Heterogeneity
To evaluate statistical heterogeneity across final studies, the Q-Cochrane test and I2 were used. The interpretation of the I2 values was as: mild (0–39.9), moderate (40–69.9), severe (70–89.9), and highly severe (90–100) [14].
Assessment of Publication Bias
To evaluate publication bias (a small study effect), Begg and Egger test and the funnel plot method were employed. A p value of < 0.01 for Begg and Egger test indicated a considerable publication bias. Also, the ‘trim and fill’ method was used to test the effect of publication bias [34–36].
Sensitivity Analysis
Sensitivity analysis was carried out via the one-out remove method to identify the effect of each study on the summary measure [37].
Risk of Bias Assessment
The risk of bias was assessed by two researchers (T.F. and R.T.) independently using the RoB tool (Cochrane risk of bias tool for randomized trials) [38]. Six items (random sequence generation, allocation concealment, blinding of participants and researchers, blinding of outcome assessment, incomplete outcome data, and selective outcome reporting) were evaluated. We set the baseline similarity of important indicators as the seventh item (online Appendix 3). Because of the small number of studies included in the high-quality group, the final score of quality assessment was calculated based on the score of primary studies on four important items: random sequence generation, blinding of participants and researchers, incomplete outcome data, and baseline similarity of important indicators. The studies that received scores for three or four items were classified as high quality, those receiving scores for two items as moderate quality, and those receiving scores for less than two items as low quality. Any disagreement between the two researchers was resolved through discussion, and, if required, a third expert was consulted to make the final decision.
Certainty of Evidence
The certainty of the evidence was assessed using the GRADE in two primary outcomes (speech perception threshold and speech-in-noise test score). The downgrading process of studies involved four factors: risk of bias (methodological quality), inconsistency (heterogeneity between trials findings), imprecision (95% Confidence interval), and publication bias, while the upgrading process included three factors: effect size, dose–response, and confounding (confounder adjusting) [39].
Results
Overview
Figure 1 shows the search results. After removing duplicates, 37,380 papers were obtained from different databases and were screened according to the inclusion criteria. A total number of 37,316 papers were removed (for the absence of a speech-in-noise test, participants’ age, auditory training method, study design, etc.). At this stage, 64 papers met the inclusion criteria. Based on the full text, some of these were excluded because of no response to emails requesting the necessary details [40–42], the absence of a speech-in-noise test, an auditory training method, or a no control group. Ultimately, 23 papers were included in the systematic review and meta-analysis. The level of inter-rater agreement in the selection process was κ = 0.16.
Fig. 1.
PRISMA flowchart of the study
Participants’ Characteristics
Overall, 859 participants were investigated in these 23 papers. The age range was 22–79 years with a mean and standard deviation of 57 (19). The participants’ hearing level was normal in 11 studies, with mild hearing loss in six, and moderate and above in six. 71 individuals used cochlear implant [11, 18, 43]. In all the included studies, both male and female participants were enrolled (Table 1).
Table 1.
Study characteristics
| Author, type of study | Sample size (female; male) | Participant characteristics | Intervention (days) (lab/home) | Outcome measures (item with the italic font was selected for this study) | |
|---|---|---|---|---|---|
| Yusof et al. [16], RCT | 76 (57; 19) |
Age (SD) Intervention: Comparison: |
Hearing status | Auditory–cognitive training (24) Lab | The Malay Hearing in Noise Test And |
|
67.6 (4.5) 65.8 (3.6) |
Normal to slight | ||||
| Chari et al. [18], cohort | 18 (8; 10) |
60.8 (16.1) 64.2 (2.3) |
Moderate and above | Melodic contour identification (20) (home) | BKB-SIN (Mean SNR in dB) |
| Rao et al. [46], CT | 22 (10; 12) |
69 (7.3) 65 (8.9) |
Mild hearing loss | RMQ (20) (home) | HINT (Mean SNR in dB) |
| Saunders et al. [17], RCT | 279 |
67.6 (7.4) 71.0 (7.5) |
Moderate and above | LACE (20) (home) |
Word in Noise Test (Mean SNR in dB) APHAB HHI WIN |
| Abrams et al. [9], RCT | 29 (10; 19) |
65.6 (5.53) 61.8 (8.43) |
Mild hearing loss | RMQ (15) (home) | HINT (SNR in dB) |
| Lidestam et al. [21], CT | 60 (33; 27) |
23.5 (2.4) : 22.7 (2.4) |
Normal to slight | Speech identification training (1) (lab) | HINT (Mean SNR in dB) |
| Ferguson et al. [15], RCT | 44 (15; 29) |
50–74 50–74 |
Mild hearing loss | Phoneme training package (24) (home) |
Adaptive Sentence List test (SRT in dB) The Glasgow hearing aid benefit profile (mean of percentage) Speech, spatial, and qualities of hearing (mean of score) |
| Anderson et al. [44], CT | 67 (39; 28) |
62.5 (3.2) 63.6 (4.1) |
Normal to slight | Brain fitness cognitive training (40) (home) | Quick-SIN (Mean SNR in dB) |
| Anderson et al. [45], RCT | 58 (35; 23) |
64.1 (5.78) 64.07 (5.22) |
Normal to slight | Brain Fitness cognitive training (40) (home) | Quick-SIN (Mean SNR in dB) |
| Olson et al. [10], RCT | 29 (13; 16) |
66 (9.6) 66 (9.3) |
Mild hearing loss | LACE (20) (home) | Quick-SIN (Mean SNR in dB) |
| Song et al. [23], CT | 60 (38; 22) |
26.0 (3.8) 23.7 (2.8) |
Normal to slight | LACE (20) (home) |
HINT (Mean SNR in dB) Quick-SIN (Mean SNR in dB) |
| Kim et al. [22], CT | 14 (3; 11) |
22.8 (2.6) 22.8 (2.6) |
Normal to slight | Spectral resolution training (4) (lab) | Korean-based speech perception in babble/white noise test (Mean SNR in dB) |
| Heidari et al. [47], CT | 32 (15; 17) |
67.56 (5.68) 70.25 (6.84) |
Normal to slight | Vowel auditory training (15) (lab) | Persian version of temporal resolution test (score) |
| Fostick et al. [24], CT | 52 (38; 14) |
65.45 (4.08) 61.33 (8.51) |
Normal to slight | Auditory temporal processing (14) (home) | Boothroyd words test in narrowband/wideband (score) General Self-Efficacy Scale (mean of score) |
| Lee et al. [12], CT | 17 (8; 9) |
79.50 (6.83) 75.71 (8.59) |
Moderate and above | Auditory Training Using Video Clips (8) (lab) | Korean matrix test (score) |
| Humes et al. [19], RCT | 45 |
71.9 (6.1) 72.0 (7.1) |
Mild hearing loss | Auditory Training Using 1. frequently occurring (American) English words (15) (home) |
Connected Speech test (score) CID (percentage) PHAP, HHIE, |
| Kim et al. [13], CT | 15 (7; 8) |
44.5 (11.31) 59.1 (19.63) |
Hearing-impaired | Story-Based Auditory Training (10) (lab) | Sentence recognition score in noise, K-IOI-HA (mean of score) |
| Rishiq et al. [26], CT | 24 (6; 18) |
68 (8.4) 69.9 (10.5) |
Mild hearing loss | ReadMyQuips [RMQ] (20) (home) | Multimodal Lexical Sentence Test for Adults (score) |
| Schumann et al. [11], CT | 27 (18, 9) |
60 (10.1) 61 (17.9) |
Moderate and above | Nonsense-syllable training (6) (lab) | The Goettingen sentence test (score) |
| Schwartz et al. [27], RCT | 27 (14; 13) |
21.6 (2.1) 21.4 (2.3) |
Normal to slight | Trivia game (12) (home) | Quick-SIN (score) |
| Krul et al. [20], CT | 30 |
18–25 18–25 |
Normal to slight | Talker-identification training (4) (lab) | Sentence recognition in noise (score) |
| Kuchinsky et al. [25], CT | 29 (12; 17) |
70.2 (8.5) 70.2 (8.5) |
Normal to slight | Speech-perception training (20) (lab) | Word stimuli in noise (score) |
| Miller et al. [43], CT | 28 (12; 16) |
75.5 (21.9) 77.75 (9.53) |
Moderate and above | Speech perception training (12) (lab) | HINT (score) |
BKB-SIN Bamford–Kowal–Bench speech-in-noise test, HINT hearing in noise test, RMQ ReadMyQuips, LACE listening and communication enhancement, WIN word in noise test, Quick SIN quick speech in noise test, Cid central institute for the deaf sentence materials, APHAB the abbreviated profile of hearing aid performance, HHI hearing handicap inventory, PHAP profile of hearing aid performance, HHIE hearing handicap inventory for the elderly, K-IOI-HA Korean international outcome inventory for hearing aids
Methodological Considerations and Outcome Measurement
Among the papers included, 16 were clinical trials, six were randomized clinical trials, and one was a cohort study. The computer-based auditory training used in these studies were listening and communication enhancement (LACE-DVD), readMyQuips, trivia game, brain fitness, speech perception assessment and training system (SPATS), phoneme/vowel/word/phrase and sentences training, talker/speech/melodic contour identification training, story-based auditory training, use of vocoded speech, etc. The comparison group included participants who had not received any auditory training, those who only used a hearing aid or those who listened to an audio file. The duration of training was one to 40 days (17 days on average). In 13 studies, the auditory training was performed at home, and in 10, it was lab-based. Outcome measurements were performed with the speech-in-noise tests. Among them, 11 reported the speech-in-noise test score, and 12 reported the speech perception threshold (Table 1).
The Effect of Intervention on Speech Perception Threshold
Among the 23 papers included in the quantitative analysis, 12 reported speech perception thresholds. In nine studies, the threshold was obtained directly from the text [8–10, 15, 16, 18, 22, 44, 45]. In two studies, the threshold was extracted from a figure via the web plot digitizer [21, 23], and the corresponding authors were asked to obtain the data for the last study in this subgroup [46].
The findings of this meta-analysis showed that computer-based auditory training could improve the speech perception threshold in subjects assigned to the training group compared to the control group (dppc2: −0.69, 95%CI: −1.11 to −0.26; I2 = 69.6%, p = 0.00), whereas this improvement on speech perception threshold was moderate (Fig. 2A). In order to evaluate the heterogeneity and to understand the effects of factors on the estimated effect size, several subgroup analyses were carried out. Analysis of the quality subgroup revealed that medium (dppc2: −1.23, 95%CI: −1.84 to −0.62) and low-quality (dppc2: −0.90, 95%CI: −1.56 to −0.24) studies showed more improvement in speech perception threshold compared to high quality (dppc2: −0.01, 95%CI: −0.38 to 0.35) studies. Analysis of the study type subgroup showed that the single cohort study (dppc2: −1.64, 95%CI: −3.25 to −0.03) and clinical trial studies (dppc2: −0.76, 95%CI: −1.04 to −0.48) had greater improvement in speech perception threshold compared to randomized clinical trials (dppc2: −0.09, 95%CI: −0.45 to 0.27). In other prespecified and post hoc subgroup analyses, no other factor was found as a potent heterogeneity source or responsible for changing the results (Table 2).
Fig. 2.
A Forest plot of studies with speech perception thresholds. B Forest plot of studies with speech-in-noise test score. ES = dppc2; white diamond = pooled effect size for all studies; horizontal line = 95% confidence interval; black diamond = pooled effect size for one study
Table 2.
Subgroup analysis in studies with speech perception threshold
| Potential factor | Layer | SMD (CI 95%) | No. of studies | Heterogeneity chi-2 | p value | I2 (%) | Interaction p value | ||
|---|---|---|---|---|---|---|---|---|---|
| Processing | Bottom-up and top-down | − 0.57 (− 0.96;− 0.19) | 8 | 15.80 | 0.027 | 55.7 | 0.896 | ||
| Only bottom-up | − 1.35 (− 2.77; 0.06) | 4 | 20.36 | 0.000 | 85.3 | ||||
| Speech in noise in AT | Yes | − 0.45 (− 0.78;− 0.12) | 8 | 9.21 | 0.238 | 24.0 | 0.138 | ||
| No | − 1.20 (− 2.35;− 0.05) | 4 | 24.77 | 0.000 | 87.9 | ||||
| AT total week | > 4 | − 0.67 (− 1.43; 0.08) | 3 | 8.38 | 0.015 | 76.1 | 0.462 | ||
| 4 | − 0.67 (− 1.30;− 0.03) | 5 | 8.99 | 0.061 | 55.5 | ||||
| < 4 | − 0.89 (− 1.94; 0.16) | 4 | 17.27 | 0.001 | 82.6 | ||||
| AT intensity (days per week) | < 3 | − 0.29 (− 0.80; 0.22) | 2 | 0.73 | 0.393 | 0.0 | 0.294 | ||
| > 4 | − 0.80 (− 1.31;− 0.28) | 10 | 34.35 | 0.000 | 73.8 | ||||
| AT site | Home | − 0.60 (− 1.01;− 0.19) | 9 | 19.87 | 0.011 | 59.7 | 0.971 | ||
| Lab | − 1.36 (− 3.00; 0.26) | 3 | 16.31 | 0.000 | 87.7 | ||||
| Age | > 65 y | − 0.45 (− 1.06; 0.15) | 4 | 6.32 | 0.097 | 52.5 | 0.144 | ||
| 31–64 y | − 0.84 (− 1.49;− 0.20) | 4 | 7.47 | 0.058 | 59.9 | ||||
| < 30 y | − 1.44 (− 2.89; 0.00) | 3 | 14.19 | 0.001 | 85.9 | ||||
| Hearing level | Normal | − 0.88 (− 1.53;− 0.23) | 6 | 22.57 | 0.000 | 77.8 | 0.191 | ||
| Mild | − 0.47 (− 1.19; 0.24) | 4 | 7.07 | 0.070 | 57.5 | ||||
| Moderate and above | − 0.67 (− 2.15; 0.80) | 2 | 3.23 | 0.072 | 69.1 | ||||
| Hearing aid use | Yes | − 0.64 (− 1.27;− 0.02) | 5 | 8.15 | 0.086 | 50.9 | 0.566 | ||
| No | − 0.72 (− 1.32;− 0.12) | 7 | 27.70 | 0.000 | 78.3 | ||||
| Study type | RCT | − 0.09 (− 0.45; 0.27) | 4 | 1.42 | 0.701 | 0.0 | 0.006 | ||
| CT | − 0.76 (− 1.04;− 0.48) | 7 | 24.67 | 0.000 | 75.7 | ||||
| Cohort | − 1.64 (− 3.25;− 0.03) | 1 | – | – | – | ||||
| Quality score | < 1 | − 0.90 (− 1.56;− 0.24) | 6 | 18.12 | 0.003 | 72.4 | 0.001 | ||
| 2 | − 1.23 (− 1.84;− 0.62) | 2 | 1.12 | 0.291 | 10.3 | ||||
| > 3 | − 0.01 (− 0.38; 0.35) | 3 | 0.34 | 0.842 | 0.0 | ||||
| Random sequence generation | Yes | − 0.09 (− 0.45; 0.27) | 4 | 1.42 | 0.701 | 0.0 | 0.004 | ||
| No | − 0.95 (− 1.56;− 0.35) | 7 | 24.67 | 0.000 | 75.7 | ||||
| Blinding of participants and researchers | Yes | − 0.01 (− 0.38; 0.35) | 3 | 0.34 | 0.842 | 0.0 | 0.001 | ||
| No | − 0.94 (− 1.47;− 0.40) | 8 | 22.67 | 0.002 | 69.1 | ||||
| Baseline similarity | Yes | − 0.39 (− 0.93; 0.15) | 5 | 14.90 | 0.005 | 73.1 | 0.200 | ||
| No | − 0.99 (− 1.73;− 0.24) | 6 | 17.80 | 0.003 | 71.9 | ||||
| All studies | − 0.69 (− 1.11;− 0.26) | 12 | 36.18 | 0.000 | 69.6 | ||||
The item with italic font: statistically significant, Cochran’s Q statistics for between-subgroup heterogeneity used for interaction p value statistics
AT Auditory training
Considering the quality score, baseline similarity in important indicators had been observed in 50% of studies. Clear information about the blinding and concealment method and sequence generation was not reported in more than 50% of studies (Fig. 3A).
Fig. 3.
Traffic light for the risk of bias assessment A studies with a speech perception threshold and B studies with a speech-in-noise test score
The Effect of Intervention on Speech-in-Noise Test Score
Of the 23 studies included in the quantitative analysis, 11 reported speech-in-noise test scores. In five, this score was obtained directly from the main text [12, 13, 24, 43, 47], while in six, the relevant information was recorded in a figure [11, 19, 20, 25–27], and the required information was achieved through the web plot digitizer.
The pooled estimates revealed that computer-based auditory training could improve the speech-in-noise test score in subjects assigned to the training group compared to the control group (dppc2: 0.71, 95%CI: 0.38 to 1.03, I2 = 17.8%, p = 0.27). Whereas this improvement in speech-in-noise test scores was moderate (Fig. 2B).
Analysis of the subgroup showed that individuals with normal hearing (dppc2:1.0, 95%CI = 0.60–1.44) had greater improvement in speech-in-noise test scores compared to participants with mild (dppc2:0.09, 95%CI =− 0.56 to 0.72) or moderated (dppc2:0.51, 95%CI =− 0.13 to 1.15) hearing loss. And also, this effect is greater in participants that did not use hearing aids (dppc2:1.02, 95%CI = 0.60–1.44) compared to participants using hearing aids (dppc2:0.34, 95%CI =− 0.08 to 0.76) (Table 3).
Table 3.
Subgroup analysis in studies with speech-in-noise test score
| Potential factor | Layer | SMD (CI 95%) | No. of studies | Heterogeneity chi-2 | p value | I2 (%) | Interaction p value |
|---|---|---|---|---|---|---|---|
| Processing | Bottom-up and top-down | 0.63 (0.17;1.07) | 8 | 9.91 | 0.194 | 29.4 | 0.459 |
| Only bottom-up | 0.87 (0.41; 1.32) | 3 | 1.47 | 0.480 | 0.0 | ||
| Speech in noise in AT | Yes | 0.38 (− 0.05; 0.81) | 7 | 3.96 | 0.681 | 0.0 | 0.071 |
| No | 0.98 (0.58; 1.37) | 4 | 4.95 | 0.175 | 39.4 | ||
| AT total week | > 4 | 0.60 (0.21; 0.99) | 7 | 6.18 | 0.404 | 2.9 | 0.422 |
| > 4 | 0.89 (0.27; 1.49) | 4 | 5.23 | 0.149 | 43.8 | ||
| AT intensity (day per week) | < 3 | 0.67 (0.28; 1.04) | 7 | 4.75 | 0.576 | 0.0 | 0.724 |
| > 4 | 0.81 (0.38; 1.02) | 4 | 7.31 | 0.063 | 0.59 | ||
| AT site | Home | 0.49 (0.06; 0.90) | 4 | 3.08 | 0.379 | 2.7 | 0.169 |
| Lab | 0.91 (0.47; 1.35) | 7 | 7.02 | 0.319 | 14.6 | ||
| Age | > 65 y | 0.59 (0.14; 1.04) | 6 | 6.17 | 0.290 | 18.9 | 0.701 |
| 31–46 y | 0.73 (0.23;1.22) | 3 | 0.72 | 0.699 | 0.0 | ||
| < 30 y | 1.30 (− 0.37; 2.98) | 2 | 4.22 | 0.040 | 76.3 | ||
| Hearing level | Normal | 1.02 (0.60; 1.44) | 5 | 4.51 | 0.324 | 11.3 | 0.047 |
| Mild | 0.09 (− 0.56; 0.72) | 2 | 0.03 | 0.856 | 0.0 | ||
| Moderate and above | 0.51 (− 0.13; 1.15) | 3 | 1.29 | 0.524 | 0.0 | ||
| Hearing aid use | Yes | 0.34 (− 0.08; 0.76) | 6 | 2.45 | 0.784 | 0.0 | 0.025 |
| No | 1.02 (0.60; 1.44) | 5 | 4.51 | 0.342 | 11.3 | ||
| Study type | RCT | 0.33 (− 0.30; 0.95) | 2 | 0.35 | 0.522 | 0.0 | 0.197 |
| CT | 0.81 (0.43; 1.17) | 9 | 10.03 | 0.263 | 20.3 | ||
| Quality score | < 1 | 0.81 (0.43; 1.17) | 9 | 10.03 | 0.263 | 20.3 | 0.365 |
| 2 | 0.52 (− 0.38; 1.42) | 1 | – | – | – | ||
| > 3 | 0.14 (− 0.74; 1.19) | 1 | – | – | – | ||
| Random sequence generation | Yes | 0.33 (− 0.30; 0.95) | 2 | 0.35 | 0.522 | 0.0 | 0.197 |
| No | 0.81 (0.43; 1.17) | 9 | 10.03 | 0.263 | 20.3 | ||
| Blinding of participants and researchers | Yes | 0.14 (− 0.74; 1.01) | 1 | – | – | – | 0.186 |
| No | 0.78 (0.44; 1.10) | 10 | 10.37 | 0.321 | 13.2 | ||
| Baseline similarity | Yes | 0.67 (0.16; 1.16) | 4 | 3.42 | 0.331 | 12.3 | 0.853 |
| No | 0.73 (0.27; 1.18) | 7 | 8.69 | 0.192 | 31.0 | ||
| All studies | 0.71 (0.38; 1.03) | 11 | 12.16 | 0.274 | 17.8 |
The item with italic font: statistically significant
AT Auditory training
Considering the quality score, baseline similarity in important indicators was observed in more than 60% of studies. In less than 40%, clear information was provided about blinding, concealment method, and sequence generation (Fig. 3B).
Sensitivity Analysis
Sensitivity analysis was conducted using the one-out remove method on both primary outcomes: speech perception threshold and speech-in-noise test scores. None of the primary studies had a different effect on the pooled measure of studies (Fig. 4).
Fig. 4.
Sensitivity analysis. A studies with speech perception threshold, B studies with speech-in-noise test score. vertical lines are indicators for the minimum, mean, and maximum value of total effect size, respectively
Publication Bias
Although a considerable publication bias should exist based on the heterogeneity in the distribution of points in the funnel plot and the result of Begg’s and Egger’s tests in both outcomes, the trim and fill method did not show a considerable amount of bias (Fig. 5).
Fig. 5.
Funnel plots for the assessment of publication bias in primary outcome A studies with speech perception threshold, B studies with speech-in-noise test score
Certainty of Evidence
The certainty of the evidence was investigated using the GRADE in two primary outcomes: speech perception threshold and speech-in-noise test score. The strength of evidence was downgraded by low-quality studies, a wide confidence interval of effect size, and publication bias. It was also upgraded by the moderate size of the intervention effect and confounder adjustment. In summary, the strength of evidence was low in both primary outcomes (Table 4).
Table 4.
The GRADE of recommendation, assessment, development, and evaluation strength of evidence for the primary outcome
| Items | Speech-in-noise test score (dppc2 = 0.71; 95% CI = 0.38 to 1.03) (I2 = 17.8%) | Speech perception threshold (dppc2 = − 0.69; 95% CI = − 1.11 to− 0.26) (I2 = 69.6%) |
|---|---|---|
| No. of studies | 11 | 12 |
| Risk of biasa | − 2 | − 2 |
| Inconsistencyb | 0 | 0 |
| Imprecisionc | − 2 | − 2 |
| Publication biasd | − 1 | − 2 |
| Effect sizee | + 1 | + 1 |
| Dose-responses | 0 | 0 |
| Confoundingf | + 1 | + 1 |
| GRADE quality | ⊕ ⊕ ⊕ ⊕ low | ⊕ ⊕ ⊕ ⊕ low |
aRisk of bias was based on randomization and baseline similarity, downgraded one level for problem with one element and two levels for problems with two elements. bDowngraded one level as the I2 value was > 50%. cDowngraded one level due to 95% CI. dDowngraded one level due to suspected publication bias. eUpgraded one level due to large effect size. fUpgraded one level due to confounder adjusting
Discussion
The present systematic review and meta-analysis aimed to examine the effect of computer-based auditory training on speech-in-noise perception in adults. The primary outcome was assessed in two subgroups: speech perception threshold and speech-in-noise test score. The effects of computer-based auditory training on quality of life, hearing disability, and heterogeneity assessments were the secondary outcomes.
The Effect of Intervention on Speech Perception Threshold
Findings from the eligible studies in this subgroup showed that computer-based auditory training can be moderately effective for speech perception thresholds in adults. However, regarding the relationship between the methodological quality of studies and the intervention results, it seems that methodological factors such as random sequence generation, blinding of participants and researchers, incomplete outcome data, and baseline similarity of important indicators are important in the validity of findings.
The Effect of Intervention on the Speech-in-Noise Test Score
In addition to the summary measure, the effect size was estimated for all but one eligible study [43], which showed a significant increase in the speech-in-noise test score following the computer-based auditory training. In this subgroup, computer-based auditory training was more effective in two groups of individuals: those with normal hearing, participants using hearing aids, and those with moderate to severe hearing loss. In other words, the participants’ hearing levels changed the efficacy of auditory training. Due to the low quality and low number of eligible studies, the results were not conclusive.
These results are in line with the previous systematic reviews. Henshaw and Ferguson noted that individual computer-based auditory training was an effective intervention for those with hearing loss. They also reported the generalization of on-task learning to improvements in speech intelligibility, cognition, and self-report of hearing. Meanwhile, because of the high methodological heterogeneity of primary studies, it was not feasible to conduct a meta-analysis and also the evidence was not robust and consistent. [6].
In line with our study, in a meta-analysis conducted by Chisholm and Arnold, it was observed that clinician-based auditory training was an effective intervention for speech recognition (Cohen’s d = 0.35), whereas the effect was small [48].
We found some methodological factors in primary studies that may have affected the efficacy of the auditory training, some of which are described below:
Speech-in-noise tests: since some primary studies attributed the small effect size of auditory training to a selection of unsuitable speech tests, it seems some tests have no required sensitivity and specificity to represent the impact of the auditory training [15–17, 26]. Accordingly, using more than one standard speech test could be beneficial. Thereafter, the researchers of secondary studies had also no difficulty selecting the desired outcome for a quantitative combination of data in the meta-analysis.
Duration of the Auditory Training
Although the effect size of the auditory training did not change significantly by the duration of training in the present study, some researchers reported that lengthening the auditory training is likely to change the results [9, 26]. Thus, it seems that in the auditory training, the minimum time needed to introduce a significant change in the speech-in-noise perception has not been yet determined.
Participants’ Commitment in the Follow-up of Auditory Training Sessions
The computer-based auditory training is not presented by specialists, and individuals perform the program at home or in a lab using their PC. Thus, reaching certainty about the completion of all training sessions is important. Olson et al. suggested this point to be an influential factor in the outcomes [10].
The Training and Assessment Site
In a study by Humes et al., the auditory training was presented at home, while the outcomes were assessed in a laboratory. According to the “encoding-specificity principle” [49], recall is improved when external cues at the time of training are present at the time of testing, based on this principle The researchers reported that different places of training and tests could affect the outcome measure [19].
Quality of Primary Studies and Strength of Evidence
Quality assessment showed that most of included studies offer low-quality evidence therefore the quality of primary studies was the most important cause of heterogeneity. Of the 23 included studies, only one study had the highest level of quality [17]. We found that clear information about blinding, concealment, and sequence generation was not present in most of the primary information. However, the outcome reporting was not selective in all primary studies. According to GRADE, two factors of wide 95% CI and low quality of primary studies were the most influential in the downgrading of evidence, and therefore the strength of evidence in the current study was low.
Some strengths of our meta-analysis should be noted. Based on our knowledge, of the meta-analyses performed until now, this meta-analysis was the most comprehensive search with more database reviewing and without setting a language limitation on primary studies.
The present study had some limitations which should be mentioned: the first and most important factor is the lack of high-quality primary studies in this area. Admittedly, the best type of trial is a randomized controlled clinical trial. Second, The lack of details on the quality of studies. Likewise, no information was available about the rate of attrition, the reason for attrition, etc. in each phase of outcome assessment.
Finally, based on the primary studies included in this systematic review and meta-analysis, computer-based auditory training would lead to improved speech-in-noise perception in adults. However, due to the low quality of primary studies and the low certainty of evidence, the results are not definite.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
We would like to thank Dr. Abbas Keshtkar for his comments on the all stage of this paper http://researchware.org/.
Abbreviation
- dppc2
Difference pretest–posttest-control
Funding
There was no external funding.
Availability of Data and Materials
The data that support the findings of this study are available from the corresponding author upon reasonable request.
Declarations
Conflict of interest
The authors have no conflicts of interest to declare.
Ethical Approval
The authors certify that this work was completed in compliance with ethical standards.
Footnotes
Mean pre-post change in the treatment group minus the mean pre-post change in the control group, divided by the pooled pretest standard deviation.
Standardized mean difference.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Deafness and hearing loss. https://www.who.int/news-room/fact-sheets/detail/deafness-and-hearing-loss. Accessed 15 Jul 2022
- 2.Shukor NFA, Lee J, Seo YJ, Han W. Efficacy of music training in hearing aid and cochlear implant users: a systematic review and meta-analysis. Clin Exp Otorhinolaryngol. 2020 doi: 10.21053/ceo.2020.00101. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Recanzone G. The effects of aging on auditory cortical function. Hear Res. 2018;366:99–105. doi: 10.1016/j.heares.2018.05.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Yeend I, Beach EF, Sharma M. Working memory and extended high-frequency hearing in adults. Ear Hear. 2019;40:458–467. doi: 10.1097/AUD.0000000000000640. [DOI] [PubMed] [Google Scholar]
- 5.Spankovich C, Gonzalez VB, Su D, Bishop CE. Self reported hearing difficulty, tinnitus, and normal audiometric thresholds, the national health and nutrition examination survey 1999–2002. Hear Res. 2018;358:30–36. doi: 10.1016/j.heares.2017.12.001. [DOI] [PubMed] [Google Scholar]
- 6.Henshaw H, Ferguson MA. Efficacy of individual computer-based auditory training for people with hearing loss: a systematic review of the evidence. PLoS ONE. 2013 doi: 10.1371/journal.pone.0062836. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Sweetow R, Palmer CV. Efficacy of individual auditory training in adults: a systematic review of the evidence. J Am Acad Audiol. 2005 doi: 10.3766/jaaa.16.7.9. [DOI] [PubMed] [Google Scholar]
- 8.Smith SL, Saunders GH, Chisolm TH, et al. Examination of individual differences in outcomes from a randomized controlled clinical trial comparing formal and informal individual auditory training programs. J Speech Lang Hear Res. 2016;59:876–886. doi: 10.1044/2016_JSLHR-H-15-0162. [DOI] [PubMed] [Google Scholar]
- 9.Abrams HB, Bock K, Ireya RL. Can a remotely delivered auditory training program improve speech-in-noise understanding? Am J Audiol. 2015;24:333–337. doi: 10.1044/2015\_AJA-15-0002. [DOI] [PubMed] [Google Scholar]
- 10.Olson AD, Preminger JE, Shinn JB. The effect of LACE DVD training in new and experienced hearing aid users. J Am Acad Audiol. 2013;24:214–230. doi: 10.3766/jaaa.24.3.7. [DOI] [PubMed] [Google Scholar]
- 11.Schumann A, Serman M, Gefeller O, Hoppe U. Computer-based auditory phoneme discrimination training improves speech recognition in noise in experienced adult cochlear implant listeners. Int J Audiol. 2015;54:190–198. doi: 10.3109/14992027.2014.969409. [DOI] [PubMed] [Google Scholar]
- 12.Lee M, Bahng J. Efficacy of auditory training using video clips for older adults who wear hearing aids. Audiol Speech Res. 2020;16:206–216 . doi: 10.21848/asr.200031. [DOI] [Google Scholar]
- 13.Kim NK, Bahng J. Development of a story based auditory training tool and evaluation of the training efficacy for adult hearing impaired listeners. Audiol Speech Res. 2017;13:133–140. doi: 10.21848/asr.2017.13.2.133. [DOI] [Google Scholar]
- 14.Higgins JPT. Cochrane handbook for systematic reviews of interventions. Hoboken: Wiley; 2019. [Google Scholar]
- 15.Ferguson MA, Henshaw H, Clark DPA, Moore DR. Benefits of phoneme discrimination training in a randomized controlled trial of 50- to 74-year-olds with mild hearing loss. Ear Hear. 2014 doi: 10.1097/AUD.0000000000000020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Yusof Y, Mukari SZ-MS, Dzulkifli MA, et al. Efficacy of a newly developed auditory–cognitive training system on speech recognition, central auditory processing and cognitive ability among older adults with normal cognition and with neurocognitive impairment. Geriatr Gerontol Int. 2019;19:768–773. doi: 10.1111/ggi.13710. [DOI] [PubMed] [Google Scholar]
- 17.Saunders GH, Smith SL, Chisolm TH, et al. A randomized control trial: supplementing hearing aid use with listening and communication enhancement (LACE) auditory training. Ear Hear. 2016;37:381–396. doi: 10.1097/AUD.0000000000000283. [DOI] [PubMed] [Google Scholar]
- 18.Chari DA, Barrett KC, Patel AD, et al. Impact of auditory-motor musical training on melodic pattern recognition in cochlear implant users. Otol Neurotol. 2020;41:e422–e431. doi: 10.1097/MAO.0000000000002525. [DOI] [PubMed] [Google Scholar]
- 19.Humes LE, Skinner KG, Kinney DL, et al. Clinical effectiveness of an at-home auditory training program. Ear Hear. 2019;40:1043–1060. doi: 10.1097/AUD.0000000000000688. [DOI] [PubMed] [Google Scholar]
- 20.Krull V, Luo X, Kirk KI. Talker-identification training using simulations of binaurally combined electric and acoustic hearing: generalization to speech and emotion recognition. J Acoust Soc Am. 2012;131:3069–3078. doi: 10.1121/1.3688533. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Lidestam B, Moradi S, Pettersson R, Ricklefs T. Audiovisual training is better than auditory-only training for auditory-only speech-in-noise identification. J Acoust Soc Am. 2014;136:EL142–EL147. doi: 10.1121/1.4890200. [DOI] [PubMed] [Google Scholar]
- 22.Kim HP, Han JH, Kwon SY, et al. Sensitivity enhancement of speech perception in noise by sound training: hearing loss simulation study. Biomed Eng Lett. 2011;1:137–142. doi: 10.1007/s13534-011-0022-y. [DOI] [Google Scholar]
- 23.Song JH, Skoe E, Banai K, Kraus N. Training to improve hearing speech in noise: biological mechanisms. Cereb Cortex. 2012;22:1180–1190. doi: 10.1093/cercor/bhr196. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Fostick L, Taitelbaum-Swead R, Kreitler S, et al. Auditory training to improve speech perception and self-efficacy in aging adults. J Speech Lang Hear Res. 2020;63:1270–1281. doi: 10.1044/2019_JSLHR-19-00355. [DOI] [PubMed] [Google Scholar]
- 25.Kuchinsky SE, Ahlstrom JB, Cute SL, et al. Speech-perception training for older adults with hearing loss impacts word recognition and effort. Psychophysiology. 2014;51:1046–1057. doi: 10.1111/psyp.12242. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Rishiq D, Rao A, Koerner T, Abrams H. Can a commercially available auditory training program improve audiovisual speech performance? Am J Audiol. 2016;25:308–312. doi: 10.1044/2016_AJA-16-0017. [DOI] [PubMed] [Google Scholar]
- 27.Schwartz K, Ringleb SI, Sandberg H, et al. Development of trivia game for speech understanding in background noise. Int J Speech Lang Pathol. 2015;17:357–366. doi: 10.3109/17549507.2014.979875. [DOI] [PubMed] [Google Scholar]
- 28.Rohatgi A (2021) Webplotdigitizer: Version 4.5
- 29.Morris SB. Estimating effect sizes from pretest-posttest-control group designs. Organ Res Methods. 2008;11:364–386. doi: 10.1177/1094428106291059. [DOI] [Google Scholar]
- 30.Morris SB, DeShon RP. Combining effect size estimates in meta-analysis with repeated measures and independent-groups designs. Psychol Methods. 2002;7:105–125. doi: 10.1037/1082-989X.7.1.105. [DOI] [PubMed] [Google Scholar]
- 31.Wilson DB (n. d. ) Practical meta-analysis effect size calculator [Online calculator]
- 32.Cohen J. Statistical power analysis for the behavioral sciences. Milton Park: Routledge; 1988. [Google Scholar]
- 33.Roeser RJ, Valente M, Hosford-Dunn H. Diagnostic procedures in audiology. Audiol Diagn. 2007;13:95–97. [Google Scholar]
- 34.Egger M, Smith GD, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ. 1997;315:629–634. doi: 10.1136/bmj.315.7109.629. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Duval S, Tweedie R. A nonparametric, “Trim and Fill” method of accounting for publication bias in meta-analysis. J Am Stat Assoc. 2012;95:89–98. doi: 10.1080/01621459.2000.10473905. [DOI] [Google Scholar]
- 36.Duval S, Tweedie R. Trim and fill: a simple funnel-plot-based method of testing and adjusting for publication bias in meta-analysis. Biometrics. 2000;56:455–463. doi: 10.1111/J.0006-341X.2000.00455.X. [DOI] [PubMed] [Google Scholar]
- 37.Schwarzer G. Methods for meta-analysis in medical research. Alex J. Sutton, Keith R. Abrams, David R. Jones, Trevor A. Sheldon and Fujian Song, Wiley, Chichester, U.K., 2000. No. of pages: xvii+317. ISBN 0-471-49066-0. Stat Med. 2003;22:3112–3114. doi: 10.1002/sim.1565. [DOI] [Google Scholar]
- 38.Higgins JPT, Altman DG, Gotzsche PC, et al. The Cochrane collaboration’s tool for assessing risk of bias in randomised trials. BMJ. 2011;343:d5928–d5928. doi: 10.1136/bmj.d5928. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Guyatt GH, Oxman AD, Vist GE, et al. GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ. 2008;336:924–926. doi: 10.1136/bmj.39489.470347.AD. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Humes LE, Kinney DL, Brown SE, et al. The effects of dosage and duration of auditory training for older adults with hearing impairment. J Acoust Soc Am. 2014;136:EL224–EL230. doi: 10.1121/1.4890663. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Lengeris A, Hazan V. The effect of native vowel processing ability and frequency discrimination acuity on the phonetic training of English vowels for native speakers of Greek. J Acoust Soc Am. 2010;128:3757–3768. doi: 10.1121/1.3506351. [DOI] [PubMed] [Google Scholar]
- 42.Sabes JH, Sweetow RW. Variables predicting outcomes on listening and communication enhancement (LACE (TM)) training. Int J Audiol. 2007;46:374–383. doi: 10.1080/14992020701297565. [DOI] [PubMed] [Google Scholar]
- 43.Miller JD, Watson CS, Kistler DJ, et al (2007) Preliminary evaluation of the speech perception assessment and training system (SPATS) with hearing-aid and cochlear-implant users. pp 050004–050004 [DOI] [PMC free article] [PubMed]
- 44.Anderson S, White-Schwoch T, Parbery-Clark A, Kraus N. Reversal of age-related neural timing delays with training. Proc Natl Acad Sci USA. 2013;110:4357–4362. doi: 10.1073/pnas.1213555110. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Anderson S, White-Schwoch T, Choi HJ, Kraus N. Training changes processing of speech cues in older adults with hearing loss. Front Syst Neurosci. 2013 doi: 10.3389/fnsys.2013.00097. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Rao A, Rishiq D, Yu L, et al. Neural correlates of selective attention with hearing aid use followed by ReadMyQuips auditory training program. Ear Hear. 2017;38:28–41. doi: 10.1097/AUD.0000000000000348. [DOI] [PubMed] [Google Scholar]
- 47.Heidari A, Moossavi A, Yadegari F, et al. Effect of vowel auditory training on the speech-in-noise perception among older adults with normal hearing. Iran J Otorhinolaryngol. 2020;32:229–236. doi: 10.22038/ijorl.2019.33433.2110. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Chisolm TH, Arnold ML. Evidence about the effectiveness of aural rehabilitation programs for adults. San Diego: Plural Publishinh, Inc.; 2012. p. 237. [Google Scholar]
- 49.Tulving E, Thomson DM. Encoding specificity and retrieval processes in episodic memory. Psychol Rev. 1973;80:352–373. doi: 10.1037/h0020071. [DOI] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.