BACKGROUND:The comprehension and production abilities of individuals with aphasia are evaluated by means of linguistic tests normed to identify patterns of strengths and difficulties. There are no culturally relevant standardised tests in Ghana to evaluate language in Akan-speaking individuals with aphasia. AIM:The aim of this study was to develop and pilot a novel culturally and linguistically relevant Akan Aphasia Language Tasks (AALT) to determine their feasibility for conducting receptive and expressive language assessments for aphasia research purposes in Akan. METHOD:AALT was designed and administered to 15 individuals with varying profiles of aphasia (nine fluent and six non-fluent) and 15 non-brain-damaged controls matched on age and sex. The tasks comprised single word comprehension for verbs, single word comprehension for nouns, sentence comprehension, object naming and action naming tasks in Akan. Stimuli were images designed using generative AI. RESULTS:All tasks showed significant differences between individuals with aphasia and non-brain-damaged controls. However, group comparisons based on aphasia type (fluent and non-fluent), showed no significant differences in performance on the noun comprehension, verb comprehension, sentence comprehension and object naming. The only statistically significant difference was on the action naming task. Overall, individuals with aphasia demonstrated greater difficulty in their comprehension of single word verbs than nouns, and comprehension of sentences. In the production tasks, individuals with non-fluent aphasia demonstrated greater difficulty in naming actions than objects. CONCLUSION:This study successfully piloted culturally and linguistically relevant aphasia assessment tasks for Akan research. The tasks proved feasible and culturally appropriate for receptive and expressive aphasia language assessment. The pilot study represents a foundational step in the development of aphasia assessments in the Akan context. The availability of this newly developed assessment tasks in Akan bridges the gap in the lack of culturally and linguistically relevant tools in Akan research for the Akan-speaking population.
Background Stuttering is a complex communication difference with significant psychological and social consequences. Negative attitudes toward people who stutter (PWS) often stem from misconceptions and limited awareness. This study aimed to investigate the attitudes of elementary school teachers toward stuttering and examine demographic factors associated with these attitudes using the Persian version of the Public Opinion Survey of Human Attributes–Stuttering (POSHA–S) in elementary school teachers in Zahedan, Iran. Methods This cross-sectional study included 150 randomly selected teachers who completed the POSHA-S. Their Overall Stuttering Score (OSS), Beliefs, and Self-Reactions subscores were compared with the international POSHA–S database using percentile benchmarking. Associations between subscores and demographic variables (age, gender, education level, marital status, and parental status) were analyzed using General Linear Modeling. Results The mean OSS was 14 (24th percentile) compared with 17 in the international POSHA–S database, indicating less favorable overall attitudes among teachers. Beliefs about PWS were less positive, particularly regarding etiology and personality, whereas Self-Reactions were moderately above average (58th percentile). In an exploratory analysis, a Gender × Marital Status interaction were significant predictors of Beliefs and OSS (p < 0.05), but not Self-Reaction; no demographic predictors reached significance for this subscore. Conclusion These findings reveal a discrepancy between teachers’ relatively positive Self-Reactions and their less favorable Beliefs about PWS. The exploratory Gender × Marital Status interaction observed for both Beliefs and OSS should be interpreted cautiously and warrants replication in larger, more balanced samples. Culturally tailored teacher education programs may help address misconceptions about stuttering and promote more informed attitudes toward students who stutter in similar educational contexts.
Differences in social communication may shape how autistic children use language with different interaction partners. This study examined lexical diversity during free play interactions with a friend and an acquaintance among 50 preschoolers (21 autistic and 29 typically developing). Each child participated in separate play sessions with the two partners. Overall, autistic children produced less lexically diverse language than TD children. TD children also demonstrated greater lexical diversity with friends than with acquaintances, whereas no systematic partner effect emerged for the autistic group. However, individual profiles based on total word production provided a more nuanced picture: most autistic children varied their verbal output across partners, but the direction of this difference was heterogeneous, with some producing more words with a friend and others with an acquaintance. Thus, autistic children's verbal output appeared sensitive to the interaction partner, although not in the relatively consistent direction observed among TD children. Associations between verbal abilities and lexical diversity were found only during interactions with acquaintances. These findings suggest that social context relates to children's language production in different ways across and within diagnostic groups and may inform individualized peer mediated interventions by supporting the selection of interaction partners who facilitate each child's communication, followed by practice across multiple partners to promote flexible language use and generalization.
While caregiver-mediated programs have been studied in research settings, little is known about their delivery and accessibility in real-world practice, including which populations they reach and their levels of engagement. This project evaluated the reach and implementation of one such program, More Than Words® (MTW), delivered virtually, as part of a large-scale government initiative for families in Ontario, Canada during the COVID-19 pandemic. Data from 2284 families with a young autistic child who registered for MTW were analyzed using the RE-AIM Framework to assess two domains: (a) reach, by examining demographic characteristics such as child age, social communication stage, and geographic location; and (b) implementation, by analyzing completion rates and attendance patterns. While the MTW program reached many families with autistic children of varying ages and social communication skills, participation from families living outside large population centres was lower than expected based on census data. Most families (71%) completed the program, but only 39% attended all of the program sessions, suggesting challenges to sustained participation. Further research is needed to understand barriers to engagement and how service delivery models can be adapted to better meet the diverse needs of families participating in the real world.
INTRODUCTION:With stroke incidence on the rise in Canada, research that can enhance the wellbeing of people living with aphasia (PLWA) is critical. To provide insights into the day-to-day social experiences of PLWA and how these experiences relate to communicative access, this study sought to: 1) describe PLWA's experiences with social participation, 2) identify challenges and supports to their social participation, and 3) gain their insights into improving communicative accessibility. METHODS:Five PLWA enrolled in a community aphasia group in Southern Ontario, Canada were included in this study. A series of individual, semi-structured interviews were conducted using supportive communication strategies for aphasia. Data collected from these interviews was analyzed using an iterative, collaborative qualitative analysis process. RESULTS:Three themes emerged during analysis: 1) coping with change, describing participants' personal experiences adapting to life with a communication disability, 2) barriers to social participation, identifying factors that negatively impacted their opportunities for interacting with others, and 3) facilitating social participation, pertaining to ways that participants felt PLWA's social participation could be improved. CONCLUSION:Improving public knowledge and awareness of aphasia was identified as a key facilitator for increasing social participation for participants in this study. Participants with perceived social support expressed having more opportunities for social participation than those without perceived social support. Present findings suggest that creating and maintaining both in-person and virtual opportunities for social participation that are communicatively accessible can help address the diverse needs of people living with aphasia.
Syntactic difficulties are widely reported in children with Developmental Language Disorder (DLD), yet detailed evidence from European Portuguese (EP) remains scarce. Language Sample Analysis (LSA) offers an ecologically valid approach to examining children's syntactic production; however, its application requires language-specific data and validated, multidimensional measures. This study aimed to provide a detailed syntactic comparison between children with DLD with syntactic impairment and typically developing (TD) peers acquiring EP, using a multidimensional LSA framework complemented by standardized assessments. Thirty-two children (60.22 ± 8.04 months) with DLD were individually matched to 32 TD peers (60.50 ± 5.96 months) on age, sex, and socioeconomic status. All participants produced picture-based narratives, which were segmented into C-units and coded for syntactic productivity, complexity, diversity, accuracy, and the production of specific structures. Group differences and associations among LSA measures and between standardized assessments were analyzed. Across all dimensions, children in the DLD group produced significantly shorter utterances (t(62) = 6.91, p < .001, d = 1.73), fewer multi-argument predicates, lower syntactic diversity, and less subordination than their TD peers. Accuracy differences were large (PGU: TD.96 vs. DLD .62), with error patterns dominated by omissions and structural errors. Several syntactic constructions that emerged in TD narratives were not observed in the narratives produced by children in the DLD group. LSA measures showed moderate-to-strong correlations with standardized assessment scores, supporting their convergent validity. These findings provide a detailed characterization of syntactic production in EP-speaking children with DLD with syntactic impairment, and underscore the value of multidimensional LSA for identifying clinically relevant syntactic vulnerabilities that may not be fully captured by standardized assessments.
BACKGROUND:Adolescents who stutter are at increased risk for social anxiety, yet the factors underlying this vulnerability remain unclear. This study examined the contributions of stuttering severity, parenting styles, social participation, and avoidance of speech situations to social anxiety among adolescents who stutter compared with adolescents who do not stutter. METHOD:Participants were 158 adolescents (79 who stutter and 79 who do not stutter) and their mothers. Adolescents who do not stutter were recruited as close friends or classmates of adolescents who stutter to enhance group comparability. Adolescents reported stuttering severity and avoidance of speech situations; parents reported parenting styles and social participation. Social anxiety was assessed using adolescent and parent reports. Analyses included paired-samples t tests, correlations, and regression models. RESULTS:Adolescents who stutter showed higher social anxiety than adolescents who do not stutter in both reports. Among adolescents who stutter, speech avoidance was associated with adolescent- and parent-reported social anxiety. Stuttering severity was associated with parent-reported social anxiety but was not significant beyond avoidance. Social participation was negatively associated with social anxiety but was not significant beyond avoidance. Among adolescents who do not stutter, adolescentreported social anxiety was associated with speech avoidance, whereas parentreported social anxiety was associated with authoritarian and permissive parenting styles. CONCLUSIONS:Speech avoidance emerged as the strongest correlate of social anxiety among adolescents who stutter. Interventions may benefit from targeting speech avoidance and promoting verbal engagement to help mitigate social anxiety in this population.
BACKGROUND:Children who are deaf and hard of hearing (DHH) require complex multidisciplinary care to achieve optimal developmental outcomes. OBJECTIVES:The present study aims to conduct a scoping review of research on Patient Navigation (PN) interventions designed to support DHH children and their families after identification of a hearing loss to determine the scope, models, and effectiveness of PN to improve access to care. METHODS:We searched PubMed, Web of Science, Embase, CINAHL, and PsycInfo databases. After duplicate records were reviewed, two researchers read titles and abstracts and screened the full texts. Consensus was reached among all authors for final study inclusion. Researchers independently extracted the data and performed data synthesis in tabular and narrative formats. Studies describing and evaluating PN interventions for DHH children with a confirmed hearing loss diagnosis were included, while studies describing an intervention without evaluation, targeted at adults, or conducted prior to confirmed diagnosis were excluded. RESULTS:Four studies met inclusion criteria. Three studies were conducted at large urban tertiary academic medical centers in the United States, two of the three included predominantly lower socioeconomic status participants, and all reported improved outcomes after PN intervention. A fourth study was conducted in Bangladesh and targeted adolescents eligible for hearing aids. CONCLUSIONS:There have been few studies evaluating PN interventions for DHH children after a confirmed hearing loss diagnosis. Additional work is needed to systematically evaluate the feasibility, effectiveness, and scalability of PN interventions on improving access to care and reducing inequities for DHH children.
PURPOSE:Menopause is a natural life stage in women that leads to many physiological and functional changes due to hormonal shifts. These hormonal changes affect the structure and function of the vocal folds, causing deterioration in voice quality. The voice plays a critical role in the functionality and quality of life of women, especially in occupational jobs requiring voice use. This study aims to establish the Turkish validity and reliability of the Menopausal Voice-Related Work Limitation Scale (MenoVWL) (Lã et al., 2025), a measurement tool designed to assess the impact of menopause-related voice changes on work performance. METHODS:The original version of the MenoVWL (Lã et al., 2025) was translated into Turkish. The Turkish version (MenoVWL-TR) was then administered to a cohort of 219 women in Ankara, Turkey. To evaluate the reliability and validity of the MenoVWL-TR, Cronbach's alpha and Confirmatory Factor Analysis (CFA) were employed, respectively. Additionally, the Kruskal-Wallis test was conducted to assess whether different menopausal stages had a significant effect on women's voices during their professional activities. RESULTS:A total of 219 adult women completed the MenoVWL-TR (mean age = 45.83 ± 11.01 years). Among them 63.0% was in pre-menopausal stage, 18.3% was in peri‑menopausal and 18.7% was in post-menopausal stage. CFA indices demonstrated that the MenoVWL-TR items showed a good fit with the data, supporting the scale's structural validity. To assess reliability, the Cronbach's Alpha coefficient was 0.947 indicating high reliability and supporting the reproducibility of the MenoVWL-TR. According to comparison results, different menopausal stages had a significant effect on women voices (p = 0.024). CONCLUSION:Voice changes during menopause manifest not as a general voice impairment, but rather as functional and work-related voice limitations. While these limitations are independent of professional use, they significantly increase with increased daily voice use. MenoVWL-TR, which allows for the functional assessment of voice changes due to menopause, is a valid and reliable tool for evaluating work-related voice limitations.
OBJECTIVE:Perceptual assessment constitutes a critical component of the evaluation process in deaf patients, providing essential information on speech perception abilities and informing auditory rehabilitation strategies. In the absence of standardized guidelines, existing clinical practices appear highly variable, both in terms of assessment tools and methodological approaches. The objective of this study was to characterize the assessment instruments currently employed by French-speaking speech-language therapists to evaluate the various domains of speech perception in deaf children and adults. DESIGN:A cross-sectional survey was disseminated online to French-speaking speech language therapists engaged in the care of deaf patients. Recruitment was conducted via specialized professional Facebook groups and targeted email outreach. RESULTS:A total of over 100 respondents completed the survey. Analysis of responses demonstrated substantial heterogeneity in the selection of assessment tools, encompassing both standardized instruments and practitioner-developed, unpublished tests lacking empirical validation. No consensus was identified regarding preferred measures or standardized assessment procedures for speech perception. CONCLUSIONS:The observed variability underscores the urgent need to establish a standardized, evidence-based protocol for speech perception assessment in deaf populations. Such standardization would facilitate methodological consistency, enhance clinical decision making, and improve the comparability of research and clinical outcomes.
BACKGROUND:In the local context, clinicians in speech-language pathology had completed a knowledge translation training aimed at promoting consistent terminology for labelling school-aged children speech, language, and communication disorders. Quantitative data found that despite high intentions to use clinical labels, actual implementation was limited. METHODS & PROCEDURES:The current study seeks to explore clinicians' views and perspectives on the factors that affected the uptake of using consistent terminology in practice, especially when communicating with families and caregivers. An online qualitative survey was used to collect qualitative data from clinicians (n = 97) after they implemented the recommended clinical labels in practice for a three-month pilot period in the publicly funded school-aged program. RESULTS:Results demonstrated that supports for using labels included improved communication with colleagues and families, positive family experience, clinician knowledge and confidence, as well as having access to ongoing supports and resources to support practice change. Barriers included differing opinions on labels, scope of practice, time and resource constraints for communicating labels, service structure, lack of clinician knowledge and confidence, family readiness, and limited opportunities for practice. CONCLUSIONS AND IMPLICATIONS:While clinicians recognized the advantages of using consistent terminology, systematic change might be needed to effect practice change. Implementation strategies such as ongoing education and training to support clinicians' knowledge and confidence, understanding family's needs, and ensuring compatibility with service structure and policy would be important considerations to successfully enact practice change.
PURPOSE:Developmental language disorder (DLD) is a neurodevelopmental disorder that affects speech and language processing. This study aimed to investigate how syllabic fundamental frequency (F0) contours and background noise affect Mandarin sentence recognition in children with DLD and identified cognitive-linguistic predictors of their performance. METHOD:Twenty-one children with DLD and 23 age-matched typically developing (TD) peers completed an open-set sentence recognition task crossing F0 (natural vs. flat) and listening condition (quiet vs. speech-shaped noise). Accuracy was scored at syllable and sub-syllabic levels (onset, rime, and tone). RESULTS:Results showed that flat F0 reduced sentence intelligibility for both groups, with larger effects in noise. Importantly, children with DLD performed worse than TD peers only when F0 was flattened. At the sub-syllabic level, onset, rime, and tone accuracy deteriorated under flat F0 and noise, particularly in DLD. Hierarchical regressions revealed dissociable predictors: verbal working memory (WM) explained a large and significant portion of variance in DLD (notably in the natural-F0, quiet condition), whereas nonverbal IQ predicted TD performance across adverse conditions; language measures were not significant predictors for both groups. CONCLUSIONS:These findings indicate that Mandarin-speaking children with DLD rely on intact F0 contours as a compensatory mechanism, a strategy that becomes fragile when prosodic or signal quality is degraded. Divergent cognitive pathways underscore the need for ecologically valid assessment and targeted supports.
PURPOSE:Previous research on language development differences in children who stutter (CWS) compared to their non-stuttering peers (CWNS) has yielded mixed results. Few studies have analyzed spontaneous speech samples from both groups studied over time. This study aims to examine expressive language by analyzing growth in expressive syntactic complexity. We also explored the relationships among syntactic complexity growth, recovery and persistence status, and stuttering severity. METHOD:Conversational speech samples were collected annually over 3-4 years from 100 CWS (32 female) and 85 CWNS (36 female), aged 3;0-5;11 at the study's start. These recorded samples were analyzed using the Index of Productive Syntax (IPSyn). We employed hierarchical linear regression modeling and Empirical Bayes' posterior means to assess group differences and individual growth trajectories over time. RESULTS:Overall, both groups exhibited similar growth in syntactic complexity. Among CWS, there was no notable difference in syntactic complexity growth between those who recovered from stuttering and those who persisted. In children who recovered, higher complexity of expressive syntax was significantly associated with lower stuttering severity. Additional analyses suggested that CWS tended to have lower scores on the IPSyn Verb Phrase subscale compared to CWNS. CONCLUSIONS:In this study, development in expressive syntactic complexity did not distinguish CWS from CWNS, nor did it differentiate between children who recovered from stuttering and those who persisted. However, in children who recovered, greater expressive syntax scores were linked to reduced stuttering severity, suggesting that more mature language ability may play a protective role in stuttering recovery.
OBJECTIVE:Approximately 11% of children three- to six-years of age have a speech, language, or other communication disorder that necessitates intervention from a skilled speech-language pathologist. Conditions of daily life that impact health (social determinants of health; SDOHs), including economic stability, healthcare access and quality, education, and neighborhood and social support, may play a role in a family's practical ability to procure speech-language therapy services and were the focus of the current study. METHOD:Within the health services utilization model, the role of predisposing variables and enabling variables, including the SDOHs, in speech-language service utilization were examined in a statewide dataset from 2010-11 (N = 2280). SDOHs were modeled as latent or manifest variables, and a logistic regression structural equation model was used to evaluate how the SDOHs moderated need on speech-language service utilization. RESULTS:Only one SDOH, neighborhood social cohesion, approached significance; other SDOHs were made of individual predictors, and neighborhood social cohesion moderated need in the final model. Despite not reaching significance, children with need for speech-language services who had high neighborhood social cohesion were up to almost twice as likely to receive services as those with low neighborhood social cohesion, though with a wide range. CONCLUSION:Supporting families' connection with neighbors and community supports may facilitate their ability to access speech-language services.
BACKGROUND:Multiple sclerosis results in sensorimotor and cognitive deficits. Previous studies have provided limited evidence of linguistic deficits, largely due to a lack of comprehensive assessments. This study addressed this gap by applying a Multilevel procedure of Linguistic Analysis (MLA) of discourse production in persons with MS (PwMS). METHODS:20 PwMS and 20 healthy controls, matched for age, education and sex, were administered tasks assessing cognitive skills and narrative discourse production abilities. The speech samples were analyzed using the MLA to assess both micro- and macrolinguistic skills. RESULTS:The MLA highlighted significant linguistic deficits in PwMS, including frequent phonological errors, reduced grammatical accuracy, limited lexical informativeness, and impaired discourse cohesion and coherence. Microlinguistic difficulties were found to correlate with macrolinguistic impairments, confirming their interconnected nature. Fatigue level significantly affected performance in PwMS. Interestingly, no significant correlations emerged between linguistic and cognitive measures. CONCLUSIONS:The comprehensive assessment of the narrative abilities of PwMS showed their impairments span lexical, grammatical, and discourse-level processes. These findings are discussed considering recent neuroimaging evidence about the neural underpinnings of discourse production and underscore the importance of incorporating training programs that target both micro- and macrolinguistic components of narrative production into clinical care's paths.
INTRODUCTION:Early Childhood Intervention (ECI) services in Spain support children aged 0-6 and their families, particularly those with developmental disorders. Recent ECI practice emphasizes family-centered, relational, and participatory approaches. This study explored families' perceptions of speech-language therapy for children with Rett syndrome (RTT) and Angelman syndrome (AS), aiming to evaluate perceived service quality, identify unmet needs, and assess the impact on children's communicative development. METHODS:A cross-sectional, descriptive study was conducted using a purpose-designed custom online survey. A convenience sample of 36 families was recruited through national and regional associations supporting RTT and AS. Due to the low prevalence of RTT and AS, analyses were conducted with an explicitly exploratory aim, emphasizing descriptive and relational patterns over inferential generalization. Quantitative analyses included correlation tests, paired-samples t-tests, multiple regression, and reliability analyses. RESULTS:Families' perceptions of professionals were primarily associated with the availability of resources, the adequacy of facilities, and the professional's attentiveness and empathy. No statistically significant associations were observed between socioeconomic status or parental quality of life and perceived service quality, and session frequency did not emerge as a predictor of perceived therapeutic effectiveness. Nevertheless, most families reported improvements in their children's communication and interaction abilities following intervention. These findings suggest the potential value of relational, individualized, and collaborative practices, while highlighting persistent gaps in professional availability, interdisciplinary coordination, and caregiver support. The small sample size limits generalizability, and results should be interpreted cautiously. CONCLUSIONS:Speech-language professionals may play a central and multidimensional role within ECI services for children with RTT and AS. Families emphasized the need for increased accessibility, emotional support, and integrated team-based approaches addressing communication, motor, and psychosocial development. Individualized and cooperative strategies could potentially enhance service quality and foster active family engagement. A ten-point framework is proposed to guide potential enhancement in ECI programs, support professional-family collaboration, and strengthen families' capacity to participate actively in their children's developmental care.
In this cross-sectional study with prospective diagnostic follow-up, vocalizations of infants and toddlers later diagnosed with autism spectrum disorder (ASD) and typically developing (TD) controls (aged 6-23 months) were coded frame-by-frame during a single Still-Face paradigm into four categories: directed; nondirected; speech-like; and non-speech-like vocalizations. We examined their relationship with the Language Development Quotient (LDQ). After FDR correction for multiple comparisons, two vocal patterns showed robust differences exclusively in the 18-23-month window: (1) directed vocalization frequency was significantly higher in TD versus ASD, and (2) nondirected vocalization duration was significantly longer in ASD versus TD, both with large effect sizes. Differences in the 6-10-and 11-17-month groups did not survive correction. Conclusions: The late divergence pattern-minimal differences in directed and nondirected vocalizations before 18 months, significant amplification at 18-23 months-identifies this window as a critical period for screening and early intervention in ASD, when social-communicative vocal patterns diverge most substantially from typical development. The findings suggest that directed vocalizations may serve as a possible low-cost behavioral marker distinguishing ASD from typical development, and that speech-like vocalizations show age-dependent associations with language outcomes, defining 18-23 months as a sensitive window for screening and early intervention.
PURPOSE:This study examines narrative production in autistic bilingual children who acquired English through either naturalistic (interactive) input or non-interactive (screen-based) exposure. Non-interactive language acquisition, often termed "unexpected bilingualism", is a growing but underexplored phenomenon in autism research, especially at the group level. METHOD:We analyzed English and Hebrew narratives from 54 bilingual children aged 4-12, divided into three groups: autistic children with non-interactive English exposure (ASD-NI), autistic children with naturalistic English exposure (ASD-Nat), and typically developing bilingual peers with naturalistic English exposure (TLD-Nat). Narrative samples were assessed for general narrative productivity measures (e.g., type and token counts, lexical diversity, syntactic complexity) and referential expression use (e.g., referential choice, referential ambiguity, and definiteness marking). RESULTS:No significant group differences were found in narrative productivity in either language. As for referential expressions, all groups used more lexical NPs during the character's introduction and more reduced forms (null/overt pronouns) during subsequent mentions of the character, reflecting typical discourse strategies. However, the ASD-Nat group produced more ambiguous pronouns than the TLD group, while no evidence of elevated ambiguity was observed in the ASD-NI group. No significant differences were observed in definiteness marking. Cross-language convergence was observed in ambiguous pronoun use. Higher verbal Theory of Mind was associated with greater lexical noun use and reduced referential ambiguity, linking clearer discourse marking to stronger socio-cognitive abilities. CONCLUSIONS:These findings highlight the linguistic competence of autistic children acquiring language through non-interactive screen-based input and highlight the importance of recognizing alternative pathways to bilingual development in autism.
PURPOSE:This study compares the Acoustic Voice Quality Index (AVQI-3) with machine learning (ML) models to evaluate their clinical utility for estimating voice quality. METHODS:Audio from 187 American English speakers (49 healthy, 138 with voice disorders) was rated for overall voice quality by six voice specialists. AVQI-3 and its six acoustic parameters were extracted, and these parameters were used to train 14 ML models (linear, curvilinear, nonlinear). Correlations and classification accuracy (normal-mild vs. moderate-severe) were compared against perceptual ratings as ground truth. RESULTS:AVQI-3 correlated strongly with perceptual ratings (Spearman rs=0.75), comparable to linear and curvilinear models, yet outperformed most nonlinear models. At a cutoff score of 2.52, AVQI-3 achieved the highest classification accuracy (0.92) with balanced sensitivity (0.90) and specificity (0.93). Among ML models, linear regression performed best (rs=0.77, accuracy=0.92, sensitivity=1.0, specificity=0.89), whereas nonlinear models showed reduced performance (average rs=0.74, accuracy=0.87, sensitivity=0.95, specificity=0.85). CONCLUSION:AVQI-3 is a simple, accessible index that quantifies voice quality as effectively as complex ML models. This is supported by the best-performing ML models being linear, indicating that linear combinations of acoustic measures are effective, accurate, and clinically interpretable, whereas added nonlinear complexity offered little benefit.
