EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC-ND 4.0

Beyond Proficiency: How Speech Attribute Alignment Shapes L2 Listening Comprehension

Simon Moxon, Nantana Sittirak

PASAA · 05.07.2026 · 72 · 357-379

DOI: 10.66947/pasaa.v72i1.3112 ↗

Kurz gefasst

Fachartikel in der Zeitschrift PASAA, online erschienen am 05.07.2026. Verfasst von Simon Moxon und 1 weiteren Autor:innen. Die Arbeit nennt 54 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This study employed a quantitative observational design to investigatewhether L2 learners' speaking proficiency, temporal-prosodic speechattributes, and speaker-listener temporal-prosodic alignment areassociated with performance on a L2 listening test. Speaker-listenertemporal-prosodic alignment was operationalized as the degree ofsimilarity between learners' produced speech and the temporal-prosodicattributes of the listening test input. This approach assumes that similarityin timing and pitch patterns may reduce perceptual processing demandsduring speech recognition. 91 Thai undergraduates submitted ten weeklyaudio journals using ALL-Talk before completing a listening teststructured on the TOEIC listening format. Learner speech and listeningtest input were analyzed for temporal-prosodic attributes using PRAAT.The student speaking proficiency scores were obtained using MicrosoftAzure via the ALL-Talk platform. These attributes were then comparedagainst student performance in the listening test using three analyticalapproaches: 1) one-way ANOVA and bootstrapped Spearmancorrelations to examine the relationship between speaking proficiencyand listening comprehension. 2) bootstrapped generalized linear modelsto assess the influence of specific temporal-prosodic attributes. 3)generalized linear mixed-effects models to investigate the role of speakerlistenertemporal-prosodic alignment. While L2 speaking proficiency didnot significantly predict comprehension performance, generalized linearmodelling revealed that Mean Pitch, Syllable Count, and Pause Countwere significant predictors. Contrary to speech alignment expectations,generalized linear mixed effects modelling showed that greater speakerlistenerdivergence in Articulation Rate significantly predicted higherlistening test scores, while shorter learner average pause durations relativeto the input were also associated with improved performance. Thesefindings complicate similarity-based accounts of speaker-listenertemporal prosody alignment and suggest that the speaking-listeningrelationship may be driven more by specific temporal and prosodicfeatures than by overall proficiency measures or uniform prosodicsimilarity.

Zitieren

APA

Moxon, S., & Sittirak, N. (2026). Beyond Proficiency: How Speech Attribute Alignment Shapes L2 Listening Comprehension. PASAA, 72, 357-379. https://doi.org/10.66947/pasaa.v72i1.3112

Vancouver

Moxon S, Sittirak N. Beyond Proficiency: How Speech Attribute Alignment Shapes L2 Listening Comprehension. PASAA. 2026;72:357-379. doi:10.66947/pasaa.v72i1.3112

BibTeX
@article{Moxon2026,
  author = {Moxon, Simon and Sittirak, Nantana},
  title = {Beyond Proficiency: How Speech Attribute Alignment Shapes L2 Listening Comprehension},
  journal = {PASAA},
  year = {2026},
  volume = {72},
  pages = {357--379},
  doi = {10.66947/pasaa.v72i1.3112}
}
Kostenlos starten →