Trained Listeners Research Articles

Inquiry into Teaching Listening and Speaking in High School English Classes Based on an Activity Perspective

This paper conducts an in-depth analysis of the issues present in current high school English listening and speaking classes and proposes corresponding improvements. Firstly, it identifies the challenges faced by high school English listening and speaking instruction, including inadequate emphasis on training listening and speaking skills by teachers, lack of operational and measurable course objectives, monotonous forms of activity organization, limitations in the use of listening materials and tools, weak connections between introductory segments and thematic content, failure to understand listening and speaking discourse based on thematic significance, neglect of teaching learning strategies, absence of learning activities in authentic contexts, and singular evaluation methods. Subsequently, a series of suggestions for enhancing teaching quality and efficiency are put forward, such as designing instruction around thematic contexts that integrate listening, speaking, reading, viewing, and writing skills; fostering the development of students’ core competencies, particularly language ability and cultural awareness; conducting comprehensive, interconnected, and practical learning activities; utilizing modern information technology to create a culturally rich learning environment; and implementing formative and multifaceted evaluation mechanisms. These improvement measures aim to enhance students’ listening and speaking abilities and cultivate them as lifelong learners with good listening and speaking habits.

Optimization of classifying accurate and misarticulated speech sounds for use in a gamified real-time ultrasound biofeedback system

Integrating ultrasound biofeedback therapy (UBT) into a real-time, gamified interface to provide articulatory feedback for speech remediation promotes an external focus of attention, thereby reducing the complex cognitive demands required for standard UBT. Previous studies have shown that accuracy of American English rhotic /r/ can be predicted by a single parameter, the difference between tongue dorsum and blade displacements measured by ultrasound imaging during speech production. This parameter has classified speech productions of rhotic syllables as correct versus misarticulated with a classification accuracy up to 85%. However, implementation of this classification approach into real-time gamified UBT, including both measurement timing and establishment of difficulty levels for progressive therapy, would benefit from optimization using a larger dataset. 2,450 productions of 10 distinct rhotic syllables (including prevocalic and postvocalic contexts) from 50 children, with and without misarticulations, were analyzed. For each production, ultrasound image sequences were processed by TonguePART software to acquire tongue displacement trajectories, and accuracy was judged by trained listeners using a visual analog scale. Analyses were conducted to optimize selection of the image frame for classification, determine parameter thresholds appropriate for real-time prediction of /r/ production accuracy, and integrate these thresholds into a difficulty level design for gamified UBT.

Vowel focused training improves listener transcription of dysarthric speech

Typical listeners can adjust to the speech of individuals with dysarthria through the process of perceptual learning. Past research has demonstrated that listeners improve in their recognition of both segments and connected speech produced by people with dysarthria. The mechanisms underlying learning of dysarthric speech are still uncertain, though it has been suggested that exposure allows listeners to tune into the segmental characteristics of speech. In the current study, we test this hypothesis by training listeners to identify vowels spoken by an individual with mild dysarthria. We employed a pre-test/post-test design where we first tested listeners on (1) vowel recognition and (2) transcription accuracy of connected speech, trained them, and then tested them again. We found that listeners who were trained on dysarthric vowels demonstrated greater improvements in vowel identification than those in the control condition. Likewise, the listeners who underwent training showed greater improvements in transcription accuracy than those in the control condition. However, the training advantage did not generalize to unfamiliar phrases. Overall, it appears that listeners are able to tune into the segmental characteristics of dysarthric speech after a short training session, which improves their recognition of longer connected speech.

An open dataset of connected speech in aphasia with consensus ratings of auditory-perceptual features.

Auditory-perceptual rating of connected speech in aphasia (APROCSA) involves trained listeners rating a large number of perceptual features of speech samples, and has shown promise as an approach for quantifying expressive speech and language function in individuals with aphasia. The aim of this study was to obtain consensus ratings for a diverse set of speech samples, which can then be used as training materials for learning the APROCSA system. Connected speech samples were recorded from six individuals with chronic post-stroke aphasia. A segment containing the first five minutes of participant speech was excerpted from each sample, and 27 features were rated on a five-point scale by five researchers. The researchers then discussed each feature in turn to obtain consensus ratings. Six connected speech samples are made freely available for research, education, and clinical uses. Consensus ratings are reported for each of the 27 features, for each speech sample. Discrepancies between raters were resolved through discussion, yielding consensus ratings that can be expected to be more accurate than mean ratings. The dataset will provide a useful resource for scientists, students, and clinicians to learn how to evaluate aphasic speech samples with an auditory-perceptual approach.

Real-Time Modulation Perception in Western Classical Music

The task of music listening involves an auditory scene analysis in which the listener makes judgments related to melody, harmony, and consonance or dissonance, all of which are made within the context of key or tonic region. Here we examine whether the process of tracking key region is independent of the process of tracking surface cues, and what surface cues may influence that process. To this end, highly trained, moderately trained, and untrained listeners listened to excerpts from string quartets, quintets, and sextets from the classical and romantic eras and responded when they heard a modulation. Each excerpt featured either a pivot chord modulation, a direct modulation, a common tone modulation, or no modulation. Listeners performed above chance across modulation conditions, and an interaction effect was observed for modulation type and participant training level. We also present an exploratory PCA that suggests that harmonic language and phrasing are both significant factors in guiding modulation perception, both of which merit further investigation.

Going Beyond Rote Auditory Learning: Neural Patterns of Generalized Auditory Learning.

The ability to generalize across specific experiences is vital for the recognition of new patterns, especially in speech perception considering acoustic-phonetic pattern variability. Indeed, behavioral research has demonstrated that listeners are able via a process of generalized learning to leverage their experiences of past words said by difficult-to-understand talker to improve their understanding for new words said by that talker. Here, we examine differences in neural responses to generalized versus rote learning in auditory cortical processing by training listeners to understand a novel synthetic talker. Using a pretest-posttest design with EEG, participants were trained using either (1) a large inventory of words where no words were repeated across the experiment (generalized learning) or (2) a small inventory of words where words were repeated (rote learning). Analysis of long-latency auditory evoked potentials at pretest and posttest revealed that rote and generalized learning both produced rapid changes in auditory processing, yet the nature of these changes differed. Generalized learning was marked by an amplitude reduction in the N1-P2 complex and by the presence of a late negativity wave in the auditory evoked potential following training; rote learning was marked only by temporally later scalp topography differences. The early N1-P2 change, found only for generalized learning, is consistent with an active processing account of speech perception, which proposes that the ability to rapidly adjust to the specific vocal characteristics of a new talker (for which rote learning is rare) relies on attentional mechanisms to selectively modify early auditory processing sensitivity.

TRAINING THE “EVERYDAY” LISTENER HOW TO RATE ACCENTED SPEECH

ABSTRACT Speech perception research has commonly relied on untrained listeners to provide impressionistic judgments of second language (L2) speech, and training listeners has been rare. The current study therefore explores the implications of training novice listeners (NLs) to rate accented speech. A total of 60 listeners were recruited, 30 who remained novice and 30 who underwent training. Both groups provided ratings of accentedness, comprehensibility, and rated listener comprehension using 5-point scales; however, the NLs did so with the traditional Likert scales, and the trained listeners (TLs) did so with evidence-based rubrics. Multi-faceted Rasch measurement was used to investigate differences in severity, consistency, and rating behavior between the two groups. Results showed that, when compared to NLs, TLs rated the speech criteria with fewer levels of severity, more consistency, and improved scalar precision, increasing the construct validity of their ratings. Implications suggest that a short training intervention can change the way NLs make judgments of L2 speech in addition to how they interact with the rating scale, thereby decreasing listener error and boosting the quality of speech measurements.

Single Word Intelligibility of Individuals with Parkinson's Disease in Noise: Pre-Specified Secondary Outcome Variables from a Randomized Control Trial (RCT) Comparing Two Intensive Speech Treatments (LSVT LOUD vs. LSVT ARTIC).

The majority of people with Parkinson’s disease (PD) experience both prosodic changes (reduced vocal volume, reduced pitch range) and articulatory changes (imprecise articulation) that often limit speech intelligibility and may contribute to significant declines in quality of life. We conducted a randomized control trial comparing two intensive treatments, voice (LSVT LOUD) or articulation (LSVT ARTIC) to assess single word intelligibility in the presence of background noise (babble and mall). Participants (64 PD and 20 Healthy) read words from the diagnostic rhyme test (DRT), an ANSI Standard for measuring intelligibility of speech, before and after one month (treatment or no treatment). Teams of trained listeners blindly rated the data. Speech intelligibility of words in the presence of both noise conditions improved in PD participants who had LSVT LOUD compared to the groups that had LSVT ARTIC or no treatment. Intensive speech treatment targeting prominent prosodic variables in LSVT LOUD had a positive effect on speech intelligibility at the single word level in PD.

Spearcon compression levels influence the gap in comprehension between untrained and trained listeners.

Auditory alarms in hospitals are ambiguous and do not provide enough information to support doctors and nurses' awareness of patient events. A potential alternative is the use of short segments of time-compressed speech, or spearcons. However, sometimes it might be desirable for patients to understand spearcons and sometimes not. We used reverse hierarchy theory to hypothesize that there will be a degree of compression where spearcons are intelligible for trained listeners but not for untrained listeners. In Experiment 1, spearcons were compressed to either 20% or 25% of their original duration. Their intelligibility was very high for trained participants, but also quite high for untrained participants. In Experiment 2 each word within each spearcon was compressed to a different degree based on the results of Experiment 1. This technique was effective in creating the desired difference in spearcon intelligibility between trained and untrained listeners. An implication of these results is that manipulating the degree of compression of spearcons "by word" can increase the effect of training so that untrained listeners reliably do not understand the content of the spearcons. (PsycInfo Database Record (c) 2021 APA, all rights reserved).

Emotion and expertise: how listeners with formal music training use cues to perceive emotion

Although studies of musical emotion often focus on the role of the composer and performer, the communicative process is also influenced by the listener’s musical background or experience. Given the equivocal nature of evidence regarding the effects of musical training, the role of listener expertise in conveyed musical emotion remains opaque. Here we examine emotional responses of musically trained listeners across two experiments using (1) eight measure excerpts, (2) musically resolved excerpts and compare them to responses collected from untrained listeners in Battcock and Schutz (2019). In each experiment 30 participants with six or more years of music training rated perceived emotion for 48 excerpts from Bach’s Well-Tempered Clavier (WTC) using scales of valence and arousal. Models of listener ratings predict more variance in trained vs. untrained listeners across both experiments. More importantly however, we observe a shift in cue weights related to training. Using commonality analysis and Fischer Z score comparisons as well as margin of error calculations, we show that timing and mode affect untrained listeners equally, whereas mode plays a significantly stronger role than timing for trained listeners. This is not to say the emotional messages are less well recognized by untrained listeners—simply that training appears to shift the relative weight of cues used in making evaluations. These results clarify music training’s potential impact on the specific effects of cues in conveying musical emotion.

An examination of articulatory skill in monolinguals and multilinguals: A tongue twister experiment

The main goal of this study is to determine whether articulatory differences exist between monolinguals (n = 19) and bilinguals (n = 21) through the analysis of tongue-twister production, following Goldrick & Blumstein (2006). The latter were divided into early bilinguals (consistent exposure to both of their languages before age 5, n = 8), mid bilinguals (exposure to L2 between 5 and 10 years, n = 5), late bilinguals (L2 exposure between 10 and 15 years, n = 3) and trilinguals (n = 5).The stimuli comprised 64 sequences that each contained four syllables (e.g., kif tif tif kif) and had to be repeated three times to a beat of a metronome (150 beats per minute). The recordings were rated by a trained listener, who gave a score of 1 for each accurately produced onset and coda consonant, and a score of 0 otherwise. In addition, a subset of the sounds produced were manually aligned in Praat to validate the raters' scoring and determine whether speakers also differed in less perceptible aspects of speech. The results show that while there are no overall differences in accuracy between the two groups (monolingual versus multilingual), bilinguals who were first exposed to their second language later in life (between 5 and 15) exhibit a statistically significant advantage in the articulation of tongue-twisters (Dugaillard & Spinu, 2019). Our findings underscore the importance of directly measuring bilingual language proficiency and incorporating this information to experimental design (DelMaschio & Abutalebi 2018, Sulpizio et al. 2019).

Multimodal perception of prominence in spontaneous speech: A methodological proposal using mixed models and AIC

Research on prominence perception has made use of animated agents and controlled speech in experimental settings, but these methodologies have disregarded some aspects of the acoustic and visual correlates of prominence. To overcome these limitations we propose a new methodological approach using spontaneous speech data. For this, we created a small database with extracts from a television talent show and neutralised the prominence-lending properties of the acoustic cues of prominence in the speech signal. In our pilot study twelve naïve listeners marked words for binary prominence (prominent vs. non-prominent) in two modalities, i.e. audio-only and audiovisual, under three conditions involving neutralisation of (a) fundamental frequency, (b) intensity, and (c) both fundamental frequency and intensity. Additionally, the marks of two trained listeners served as control condition. Different generalised linear mixed models were estimated and compared using the Akaike Information Criterion (AIC). The most parsimonious model was then examined using traditional null-hypothesis testing in order to provisionally establish the effects of our independent variables on prominence marking. We argue that spontaneous speech can be successfully applied to the study of the multimodal perception of prominence.

From categories to gradience: Auto-coding sociophonetic variation with random forests

The time-consuming nature of coding sociophonetic variables that are typically treated as categorical represents an impediment to addressing research questions around these variables that require large volumes of data. In this paper, we apply a machine learning method, random forest classification (Breiman, 2001), to automate coding (categorical prediction) of two English sociophonetic variables traditionally treated as categorical, non-prevocalic /r/ and word-medial intervocalic /t/, based on tokens’ acoustic signatures. We found good performance for binary classifiers of non-prevocalic /r/ (Absent versus Present) and medial /t/ (Voiced versus Voiceless), but not for medial /t/ with a six-way coding distinction (largely due to some codes being sparsely represented in the training data). This method also yields rankings of acoustic measures in terms of importance in classification. Beyond any individual measures, this method generates probabilistic predictions of variation (classifier probabilities) that represent a composite of the acoustic cues fed into the model. In a listening experiment, we found that not only did classifier probabilities significantly capture gradience in trained listeners’ perceptions of rhoticity, they better predicted listeners’ perceptions than individual acoustic measures. This method thus represents a new approach to reconciling the categorical and continuous dimensions of sociophonetic variation.

User interactions and behaviors in a large-scale online emotional support service

Among the most important reasons why people communicate with each other is to share and support each other through emotional problems, yet most online social systems are uncomfortable or unsafe spaces for this purpose. This has led to the development of online emotional support systems, where users needing to speak to someone can anonymously connect to a crowd of trained listeners for a one-on-one conversation. Toward understanding the qualities of this emerging type of online social system, this article examines the users, conversations, and activities performed across 7 Cups, a massive, vibrant emotional support system with a community of listeners ready to help those with any number of emotional issues. The study makes intriguing insights along the worldwide adoption of the service, the need of its users to seek support from many others, a power-law effect of listener popularity, that users have a penchant to connecting to others along common interests and that a core periphery-like structure emerges among conversation networks, and identifies qualities of the system that drive user engagement and retention. We further study the words and actions of misbehaving users who have been reported on or blocked, and build a machine learning classifier able to anticipate their undesirable actions with reasonable accuracy and very low false positive rate. The qualities recovered gave insight into the user dynamics and communication structure of an online emotional support service, the features that drive engagement, and a means of identifying misbehaving users automatically.

Hypernasality in singing among children with cleft palate: a preliminary study

The aim of this study was to document differences in hypernasality during speaking and singing among children with cleft palate and to compare nasality score ratings of trained and untrained listeners. Twenty subjects with cleft palate aged between 7 and 12 years participated in this study. Audio recordings were made of the children reading a passage and singing a common local song, both in the Malay language. The degree of hypernasality was judged through perceptual assessment. Three trained listeners (a speech therapist, a classical singer, and a linguistic expert – all academicians) and two untrained listeners (a cleft volunteer worker and a national high school teacher) assessed the recordings using a visual analogue scale (VAS). Inter-rater and intra-rater reliability for hypernasality in both speaking and singing were verified using the intra-class correlation coefficient (ICC). A significant reduction in hypernasality was observed during singing as compared to speaking, indicating that hypernasality reduces when a child with cleft palate sings. The act of singing significantly reduces hypernasality. The outcome of this study suggests that children with cleft palate would benefit from singing exercises to ultimately reduce hypernasality. However, future research is needed to objectively measure nasality in singing compared to speaking.

Auditory-Perceptual Rating of Connected Speech in Aphasia.

Purpose Auditory-perceptual assessment, in which trained listeners rate a large number of perceptual features of speech samples, is the gold standard for the differential diagnosis of motor speech disorders. The goal of this study was to investigate the feasibility of applying a similar, formalized auditory-perceptual approach to the assessment of language deficits in connected speech samples from individuals with aphasia. Method Twenty-seven common features of connected speech in aphasia were defined, each of which was rated on a 5-point scale. Three experienced researchers evaluated 24 connected speech samples from the AphasiaBank database, and 12 student clinicians evaluated subsets of 8 speech samples each. We calculated interrater reliability for each group of raters and investigated the validity of the auditory-perceptual approach by comparing feature ratings to related quantitative measures derived from transcripts and clinical measures, and by examining patterns of feature co-occurrence. Results Most features were rated with good-to-excellent interrater reliability by researchers and student clinicians. Most features demonstrated strong concurrent validity with respect to quantitative connected speech measures computed from AphasiaBank transcripts and/or clinical aphasia battery subscores. Factor analysis showed that 4 underlying factors, which we labeled Paraphasia, Logopenia, Agrammatism, and Motor Speech, accounted for 79% of the variance in connected speech profiles. Examination of individual patients' factor scores revealed striking diversity among individuals classified with a given aphasia type. Conclusion Auditory-perceptual rating of connected speech in aphasia shows potential to be a comprehensive, efficient, reliable, and valid approach for characterizing connected speech in aphasia.

Cortical reorganization following auditory spatial training in listeners with sensorineural hearing impairment: A high-density electroencephalography study

The present study is intervention-based research aimed at remediation of spatial deficits in listeners with sensorineural hearing impairment (SNHI), through the use of virtual acoustic technology. A mixed group design comprising both within (pre-test, post-test control group design) and across the groups (standard group) comparisons were performed. The study included 37 participants, who were divided into three groups. Groups I and II consisted of SNHI listeners, while group III comprised normal hearing (NH) listeners. The study was conducted in three phases. At the pre-training phase, electroencephalographic (EEG) recordings were acquired from all the three groups using spatial deviants presented in P300 paradigm. Following this, group I listeners underwent virtual acoustic space training (VAST), and post-training EEG recordings were obtained. EEG recordings were also acquired from group II listeners in second evaluation without subjecting them to any formal spatial training. Results of unpaired t-tests, grand average waveforms and scalp topographies of offline processed waveforms revealed significant differences between SNHI and NH listeners. Furthermore, spatio-temporal analyses showed the emergence of new scalp maps in post-training phase in trained listeners and no topographic changes in untrained SNHI group, suggestive of benefit derived from VAST right at the fundamental level (cortical) of spatial processing.

Automated parsing of sound level meter data into artifacts, natural sources, and noise

The National Park Service (NPS) has nearly one million hours of 1/3rd octave band sound level measurements collected at hundreds of sites throughout the system. One of the principal purposes of these data is to measure the background or residual sound level against which all transient sounds are heard. Historically this analysis has depended upon trained listeners to identify which sound sources can be heard in segments of these data. These annotated records were then used to calculate an adjusted median sound level. To pursue a more efficient, automated method to deliver this result, matrix decomposition methods were tested to characterize their capacity to model time series of sound level spectra as low-rank decompositions, including options to account for anomalous events. These methods were combined with sound source identification based on component spectral properties and time weightings to yield promising approaches for automating NPS analyses. These results may also offer options for detecting and removing the effects of pseudonoise due to the turbulence generated by air flowing past the microphone.

Sound Distortion Errors in Aphasia With Apraxia of Speech.

Purpose Loosely defined diagnostic criteria for acquired apraxia of speech (AOS) limit clinicians' ability to diagnose the disorder validly and reliably. The purpose of this study was to contribute to the development of more precise diagnostic guidelines by characterizing the frequency and quality of sound distortion errors in speakers with clinically diagnosed AOS. Method Audio-recorded motor speech evaluations from 24 speakers with AOS and aphasia were analyzed by trained listeners using a narrow phonetic transcription protocol that included 12 distortion categories. We calculated percentage of segments transcribed with phonemic error, distortion error, and a combination of phonemic and distortion error. Results Distortion frequency varied substantially across participants, distributing on a continuum from 5% to 22% of segments. The frequency of phonemic errors was significantly greater than the frequency of distortion errors, which, in turn, was greater than the frequency of distorted substitution errors. The most common distortion qualities were voicing ambiguity and segment lengthening, but over 40% of distortion errors were distributed across an assortment of tongue modifications. Conclusions The results replicated observations from previous studies of speakers with quantitatively defined AOS in a new sample of participants with clinically diagnosed AOS. Similar distortion qualities were observed across studies, offering focus for diagnosticians and guidance for operationalizing future measures. The broad performance continua we observed help explain why binary classification of the presence/absence of AOS can be challenging and indicate a need to develop quantitative norms.

Can meditation improve attention in older adults? Study protocol for a 4-week proof-of-concept intervention

BackgroundFalls are a major health care concern for our aging population. Previous research has identified impaired sustained attention as a risk factor for falls. Recently, meditation has been shown to improve different types of attention in various populations. However, there are no studies to date examining whether meditation training can improve sustained attention and mobility in older adults.MethodsWe are conducting a 4-week proof-of-concept meditation intervention. We will recruit community-dwelling older adults. Participants will be randomized into one of two groups: (1) meditation training or (2) music listening (control). All participants will complete three 20-min group sessions per week and will be encouraged to continue their practice independently on the remaining days each week. Our primary outcome measure is behavioral performance on the Sustained Attention to Response Task (SART). Our secondary and tertiary outcomes include electroencephalograms (EEG) to assess attention and cognitive processing, mobility, and executive function.DiscussionOur proof-of-concept intervention aims to examine whether meditation training can improve sustained attention in older adults, who are known to be susceptible to falls. Importantly, our research has the potential to inform future clinical trials aimed at improving mobility and reducing falls risk in our aging population.Trial registrationClinicaltrials.gov ID NCT03417635.