This study examined the cross‐linguistic shareability of word recognition skills with a partial least squares structural equation model. A total of 199 Chinese‐speaking English learners were measured on their first (L1) and second language (L2) word form analysis skills (orthographic, phonological, and morphological processing skills) and L2 word recognition. The analysis demonstrated that (a) at the construct level, these L1 word form analysis skills systematically transferred to the corresponding L2 skills; (b) however, at the outcome level, their transfer to L2 word recognition differed: L1 phonological processing skills affected L2 word recognition directly and indirectly via L2 phonological processing skills, and L1 orthographic and morphological processing only indirectly affected L2 word recognition. We argued that there were fundamental differences in the cross‐linguistic shareability and thus transfer patterns across the word form analysis skills. These differences are crucial to our understanding of reading universals and L2 reading pedagogy.
{"title":"Cross‐Linguistic Shareability of Word Recognition Skills Among Chinese‐Speaking English Learners: A Partial Least Squares Structural Equation Model","authors":"Xiaomeng Li, Keiko Koda","doi":"10.1111/lang.70014","DOIUrl":"https://doi.org/10.1111/lang.70014","url":null,"abstract":"This study examined the cross‐linguistic shareability of word recognition skills with a partial least squares structural equation model. A total of 199 Chinese‐speaking English learners were measured on their first (L1) and second language (L2) word form analysis skills (orthographic, phonological, and morphological processing skills) and L2 word recognition. The analysis demonstrated that (a) at the construct level, these L1 word form analysis skills systematically transferred to the corresponding L2 skills; (b) however, at the outcome level, their transfer to L2 word recognition differed: L1 phonological processing skills affected L2 word recognition directly and indirectly via L2 phonological processing skills, and L1 orthographic and morphological processing only indirectly affected L2 word recognition. We argued that there were fundamental differences in the cross‐linguistic shareability and thus transfer patterns across the word form analysis skills. These differences are crucial to our understanding of reading universals and L2 reading pedagogy.","PeriodicalId":51371,"journal":{"name":"Language Learning","volume":"53 1","pages":""},"PeriodicalIF":4.4,"publicationDate":"2025-10-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145382196","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"文学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Studies have explored the relationship between text readability and processing effort in second language (L2) reading—as evidenced by eye movements. However, these studies generally relied on short texts, raising concerns about the validity of the analyzed data. This study reexamined these relationships using open‐source eye‐tracking data from L2 English learners who read longer passages (those over 200 words). The passages were analyzed for different readability indices and various linguistic features, which were subsequently used to predict some passage‐level eye‐tracking measures. Bayesian analysis revealed that complex linguistic features, primarily lexical sophistication, play a significant role in predicting these measures. However, the benefits of using these features were not much greater than those of using readability indices or simple linguistic features, such as word and sentence length. This study concludes that simple linguistic features can be effective predictors of processing effort in L2 text reading, considering their interpretability and low computational cost.
{"title":"Revisiting Text Readability and Processing Effort in Second Language Reading: Bayesian Analysis of Eye‐Tracking Data","authors":"Shingo Nahatame, Kazuhiro Yamaguchi","doi":"10.1111/lang.70011","DOIUrl":"https://doi.org/10.1111/lang.70011","url":null,"abstract":"Studies have explored the relationship between text readability and processing effort in second language (L2) reading—as evidenced by eye movements. However, these studies generally relied on short texts, raising concerns about the validity of the analyzed data. This study reexamined these relationships using open‐source eye‐tracking data from L2 English learners who read longer passages (those over 200 words). The passages were analyzed for different readability indices and various linguistic features, which were subsequently used to predict some passage‐level eye‐tracking measures. Bayesian analysis revealed that complex linguistic features, primarily lexical sophistication, play a significant role in predicting these measures. However, the benefits of using these features were not much greater than those of using readability indices or simple linguistic features, such as word and sentence length. This study concludes that simple linguistic features can be effective predictors of processing effort in L2 text reading, considering their interpretability and low computational cost.","PeriodicalId":51371,"journal":{"name":"Language Learning","volume":"27 1","pages":""},"PeriodicalIF":4.4,"publicationDate":"2025-10-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145397922","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"文学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The present study investigated how instruction across explicit versus implicit conditions interacts with aptitude to affect gains in the pronunciation of lexeme‐initial biconsonantal clusters among learners of English as a second language. The study further examined whether condition‐congruent aptitudes correlate with emotions that learners experience during instruction and whether and how such emotional reactions mediate the effect of condition‐congruent aptitudes on the learners’ pronunciation outcomes. Results from the general linear model analyses indicated a significant effect of instruction on pronunciation outcomes, as well as a superiority for explicit instruction compared with implicit instruction. The results further revealed a significant effect of the interaction of aptitude and instructional condition on the learners’ pronunciation gains. Condition‐congruent aptitudes were also found to correlate strongly with different emotions that learners experienced. Furthermore, the results of the mediated regression analyses in R revealed that emotions mediated the influence of condition‐congruent aptitudes on pronunciation outcomes.
{"title":"Gains in Pronunciation of Lexeme‐Initial Biconsonantal Clusters Among Second Language Learners: An Individual Differences–Treatment Interaction Design","authors":"Mohammad N. Karimi, Parisa Ashkani","doi":"10.1111/lang.70010","DOIUrl":"https://doi.org/10.1111/lang.70010","url":null,"abstract":"The present study investigated how instruction across explicit versus implicit conditions interacts with aptitude to affect gains in the pronunciation of lexeme‐initial biconsonantal clusters among learners of English as a second language. The study further examined whether condition‐congruent aptitudes correlate with emotions that learners experience during instruction and whether and how such emotional reactions mediate the effect of condition‐congruent aptitudes on the learners’ pronunciation outcomes. Results from the general linear model analyses indicated a significant effect of instruction on pronunciation outcomes, as well as a superiority for explicit instruction compared with implicit instruction. The results further revealed a significant effect of the interaction of aptitude and instructional condition on the learners’ pronunciation gains. Condition‐congruent aptitudes were also found to correlate strongly with different emotions that learners experienced. Furthermore, the results of the mediated regression analyses in R revealed that emotions mediated the influence of condition‐congruent aptitudes on pronunciation outcomes.","PeriodicalId":51371,"journal":{"name":"Language Learning","volume":"86 1","pages":""},"PeriodicalIF":4.4,"publicationDate":"2025-10-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145260694","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"文学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
A key debate in second language acquisition research revolves around the relative significance of explicit and implicit learning conditions in grammar learning. However, little is known about the potential of learners’ extramural (i.e., out‐of‐class) language use in fostering implicit and/or automatized knowledge as compared to explicit knowledge. Therefore, the present study investigated the effect of extramural English (EE) on implicit, automatized, and explicit knowledge among 13‐ to 14‐year‐old learners in Austria and Sweden (N = 213). According to the results generated by linear mixed models, EE use can predict implicit and/or automatized knowledge. The strength of this relationship seems to be influenced by the extent, starting age, and nature of EE use, with productive but also multimodal and highly interest‐driven activities showing the greatest effect. Explicit knowledge, however, was not affected by EE use. Overall, learner reports on the role of EE and instruction in grammar learning support the findings.
{"title":"Exploring the Potential of Extramural English in the Development of Implicit, Automatized, and Explicit Knowledge of Grammar","authors":"Alexandra Schurz","doi":"10.1111/lang.70008","DOIUrl":"https://doi.org/10.1111/lang.70008","url":null,"abstract":"A key debate in second language acquisition research revolves around the relative significance of explicit and implicit learning conditions in grammar learning. However, little is known about the potential of learners’ extramural (i.e., out‐of‐class) language use in fostering implicit and/or automatized knowledge as compared to explicit knowledge. Therefore, the present study investigated the effect of extramural English (EE) on implicit, automatized, and explicit knowledge among 13‐ to 14‐year‐old learners in Austria and Sweden (<jats:italic>N</jats:italic> = 213). According to the results generated by linear mixed models, EE use can predict implicit and/or automatized knowledge. The strength of this relationship seems to be influenced by the extent, starting age, and nature of EE use, with productive but also multimodal and highly interest‐driven activities showing the greatest effect. Explicit knowledge, however, was not affected by EE use. Overall, learner reports on the role of EE and instruction in grammar learning support the findings.","PeriodicalId":51371,"journal":{"name":"Language Learning","volume":"86 1","pages":""},"PeriodicalIF":4.4,"publicationDate":"2025-10-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145241295","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"文学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The current study examined how children apply their phonological knowledge to recognize translation equivalents in a foreign language. Target words for recognition were either phonologically similar (cognate) or dissimilar (noncognate) to words they already knew in their first language. To examine how recognition of such words is impacted by their position in the utterance, target words were either embedded in an edge (Experiment 1) or non‐edge (Experiment 2) position in the carrier phrase. Results show that preschool‐aged children can recognize words from a foreign language when those words are phonologically similar to words they know in their first language, and this is not constrained by the word's location within an utterance. Children are at an advantage recognizing new lexical items in a new language if those items are similar in form to words they already know in their first language.
{"title":"Children's Foreign Word Recognition at First Exposure: The Role of Phonological Similarity and Utterance Position","authors":"Katie Von Holzen, Rochelle S. Newman","doi":"10.1111/lang.70009","DOIUrl":"https://doi.org/10.1111/lang.70009","url":null,"abstract":"The current study examined how children apply their phonological knowledge to recognize translation equivalents in a foreign language. Target words for recognition were either phonologically similar (cognate) or dissimilar (noncognate) to words they already knew in their first language. To examine how recognition of such words is impacted by their position in the utterance, target words were either embedded in an edge (Experiment 1) or non‐edge (Experiment 2) position in the carrier phrase. Results show that preschool‐aged children can recognize words from a foreign language when those words are phonologically similar to words they know in their first language, and this is not constrained by the word's location within an utterance. Children are at an advantage recognizing new lexical items in a new language if those items are similar in form to words they already know in their first language.","PeriodicalId":51371,"journal":{"name":"Language Learning","volume":"17 1","pages":""},"PeriodicalIF":4.4,"publicationDate":"2025-09-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145088947","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"文学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Jonathan Malone, Bronson Hui, Nick Pandža, Tetiana Tytko
This study examined second language vocabulary processing and learning in reading only (RO) versus reading while listening (RWL). 119 English learners read or read‐while‐listening to a story embedded with 25 pseudowords, 10 times each, and had their eye movements tracked. Gaze duration (GD) and total reading time (TRT) measured processing of target vocabulary, and learning outcome tests included recognition and recall tests in both visual and auditory item modalities. Eye‐tracking analyses revealed (a) a significant decrease in GD and TRT for both groups, and (b) faster reading times and greater rate of decrease across instances in RO. All learning outcomes indicated superior scores from RWL in auditory items and item modality effects in RO, but not RWL, while a group‐by‐item‐modality interaction was uncovered on two of three outcomes. Results indicated a slightly slower multimodal reading process, with superior learning of phonological information from RWL, without detriment to orthographic learning.
{"title":"Eye Movements, Item Modality, and Multimodal Second Language Vocabulary Learning: Processing and Outcomes","authors":"Jonathan Malone, Bronson Hui, Nick Pandža, Tetiana Tytko","doi":"10.1111/lang.70007","DOIUrl":"https://doi.org/10.1111/lang.70007","url":null,"abstract":"This study examined second language vocabulary processing and learning in reading only (RO) versus reading while listening (RWL). 119 English learners read or read‐while‐listening to a story embedded with 25 pseudowords, 10 times each, and had their eye movements tracked. Gaze duration (GD) and total reading time (TRT) measured processing of target vocabulary, and learning outcome tests included recognition and recall tests in both visual and auditory item modalities. Eye‐tracking analyses revealed (a) a significant decrease in GD and TRT for both groups, and (b) faster reading times and greater rate of decrease across instances in RO. All learning outcomes indicated superior scores from RWL in auditory items and item modality effects in RO, but not RWL, while a group‐by‐item‐modality interaction was uncovered on two of three outcomes. Results indicated a slightly slower multimodal reading process, with superior learning of phonological information from RWL, without detriment to orthographic learning.","PeriodicalId":51371,"journal":{"name":"Language Learning","volume":"15 1","pages":""},"PeriodicalIF":4.4,"publicationDate":"2025-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145072538","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"文学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We investigate what is learned from exposure to usage in verbal morphology using an error correction mechanism within an associative learning framework. We computationally simulated how second language (L2) learners would respond to naturalistic input of aspectual usage, characterized by “imperfect contingencies,” given two types of instructions: grammatical versus lexical. To test these predictions, English native speakers (N = 80) completed three online training sessions in two conditions (grammatical vs. lexical) over 3 days, learning 21 Polish verbs across 189 exposures; and a 63‐item posttest on Day 4 (50% seen, 50% grammatical). The results confirmed the simulation predictions: The grammatical group performed better through stronger performance in contexts that allow only one aspect; the lexical group was slightly better in contexts where both aspects were possible. Rules offer some advantage early on, especially when the exemplars are already unambiguous, whereas an exemplar‐based approach promises a more flexible system in the longer run.
{"title":"Error Correction Learning of Second Language Verbal Morphology: Associating Imperfect Contingencies in Naturalistic Frequency Distributions","authors":"Justyna Mackiewicz, Petar Milin, Dagmar Divjak","doi":"10.1111/lang.70006","DOIUrl":"https://doi.org/10.1111/lang.70006","url":null,"abstract":"We investigate what is learned from exposure to usage in verbal morphology using an error correction mechanism within an associative learning framework. We computationally simulated how second language (L2) learners would respond to naturalistic input of aspectual usage, characterized by “imperfect contingencies,” given two types of instructions: grammatical versus lexical. To test these predictions, English native speakers (<jats:italic>N</jats:italic> = 80) completed three online training sessions in two conditions (grammatical vs. lexical) over 3 days, learning 21 Polish verbs across 189 exposures; and a 63‐item posttest on Day 4 (50% seen, 50% grammatical). The results confirmed the simulation predictions: The grammatical group performed better through stronger performance in contexts that allow only one aspect; the lexical group was slightly better in contexts where both aspects were possible. Rules offer some advantage early on, especially when the exemplars are already unambiguous, whereas an exemplar‐based approach promises a more flexible system in the longer run.","PeriodicalId":51371,"journal":{"name":"Language Learning","volume":"29 1","pages":""},"PeriodicalIF":4.4,"publicationDate":"2025-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145056719","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"文学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Kevin McManus, Kelly Bayas, Katherine Kerschen, Yulia Khoruzhaya, Jingyuan Zhuang, Alex Magnuson
We closely replicated Ellis and Sagarra (2010b), a seminal study that demonstrated clear effects of blocking in second language (L2) learning. In that study, English‐speaking learners completed different types of pretraining about Latin temporal expressions (adverbs, verbs, none) to investigate how knowledge about specific cues influenced L2 outcomes. Results showed that the type of pretraining strongly influenced interpretation and production. To date, however, no replication has fully confirmed these findings, resulting in open questions about the nature and role of blocking in L2 learning. Unlike prior studies, we addressed these concerns without modifying the initial study's materials and procedures. Our close replication did not confirm the initial study's findings due to performance following verb pretraining: Pretraining on verbs did not bias attention to verbs in subsequent use. Theoretical refinements and methodological implications of this replication study are discussed, especially roles for prior experience, competition, and the linguistic properties of cues.
{"title":"Revisiting Blocking Effects in Second Language Learning: A Close Replication of Ellis and Sagarra (2010b)","authors":"Kevin McManus, Kelly Bayas, Katherine Kerschen, Yulia Khoruzhaya, Jingyuan Zhuang, Alex Magnuson","doi":"10.1111/lang.70005","DOIUrl":"https://doi.org/10.1111/lang.70005","url":null,"abstract":"We closely replicated Ellis and Sagarra (2010b), a seminal study that demonstrated clear effects of blocking in second language (L2) learning. In that study, English‐speaking learners completed different types of pretraining about Latin temporal expressions (adverbs, verbs, none) to investigate how knowledge about specific cues influenced L2 outcomes. Results showed that the type of pretraining strongly influenced interpretation and production. To date, however, no replication has fully confirmed these findings, resulting in open questions about the nature and role of blocking in L2 learning. Unlike prior studies, we addressed these concerns without modifying the initial study's materials and procedures. Our close replication did not confirm the initial study's findings due to performance following verb pretraining: Pretraining on verbs did not bias attention to verbs in subsequent use. Theoretical refinements and methodological implications of this replication study are discussed, especially roles for prior experience, competition, and the linguistic properties of cues.","PeriodicalId":51371,"journal":{"name":"Language Learning","volume":"58 1","pages":""},"PeriodicalIF":4.4,"publicationDate":"2025-07-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144766105","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"文学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
How flexible are grammars after puberty? To answer this, we test returnees: heritage speakers (HS) born in an immigration context who returned to their homeland in later years. If returnees are targetlike, then language is still malleable after puberty; in contrast, if maturational effects are in play, postpuberty returnees will show variability. Thirty‐two Turkish‐American returnees, 30 Turkish HS in the United States, and 30 monolinguals completed an acceptability judgment task and a sentence repetition task on a vulnerable structure in Turkish HS, namely relative clauses. Results showed that (a) returnees patterned with monolinguals, who significantly outperformed HS in both tasks; and (b) age of return to Turkey, length of residence, and contact with English upon return were not significant predictors of returnees’ performance. This suggests that complex morphosyntax is malleable after puberty. These findings provide a unique angle on the roles of age and input variables in bilingual language acquisition and maintenance.
{"title":"How Flexible Are Grammars Past Puberty? The Case of Relative Clauses in Turkish‐American Returnees","authors":"Aylin Coşkun Kunduz, Silvina Montrul","doi":"10.1111/lang.70003","DOIUrl":"https://doi.org/10.1111/lang.70003","url":null,"abstract":"How flexible are grammars after puberty? To answer this, we test returnees: heritage speakers (HS) born in an immigration context who returned to their homeland in later years. If returnees are targetlike, then language is still malleable after puberty; in contrast, if maturational effects are in play, postpuberty returnees will show variability. Thirty‐two Turkish‐American returnees, 30 Turkish HS in the United States, and 30 monolinguals completed an acceptability judgment task and a sentence repetition task on a vulnerable structure in Turkish HS, namely relative clauses. Results showed that (a) returnees patterned with monolinguals, who significantly outperformed HS in both tasks; and (b) age of return to Turkey, length of residence, and contact with English upon return were not significant predictors of returnees’ performance. This suggests that complex morphosyntax is malleable after puberty. These findings provide a unique angle on the roles of age and input variables in bilingual language acquisition and maintenance.","PeriodicalId":51371,"journal":{"name":"Language Learning","volume":"33 1","pages":""},"PeriodicalIF":4.4,"publicationDate":"2025-07-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144693929","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"文学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Danwei Cai, Ben Naismith, Maria Kostromitina, Zhongwei Teng, Kevin P. Yancey, Geoffrey T. LaFlair
Globalization and increases in the numbers of English language learners have led to a growing demand for English proficiency assessments of spoken language. In this paper, we describe the development of an automatic pronunciation scorer built on state-of-the-art deep neural network models. The model is trained on a bespoke human-rated dataset that reflects current perspectives on pronunciation and intelligibility. The new scorer is evaluated along three criteria: How well it explains expert human ratings, how it compares to other state-of-the-art automatic pronunciation scorers in explaining expert human ratings, and the extent to which it exhibits bias toward different groups of test takers. Results indicate that the proposed scorer shows strong positive correlations with expert human ratings and outperforms other scorers. However, the scorer shows some bias related to audio quality and language family groups. We conclude with future directions for mitigating bias and argue that this scorer holds potential for use in operational settings.
{"title":"Developing an Automatic Pronunciation Scorer: Aligning Speech Evaluation Models and Applied Linguistics Constructs","authors":"Danwei Cai, Ben Naismith, Maria Kostromitina, Zhongwei Teng, Kevin P. Yancey, Geoffrey T. LaFlair","doi":"10.1111/lang.70000","DOIUrl":"10.1111/lang.70000","url":null,"abstract":"<p>Globalization and increases in the numbers of English language learners have led to a growing demand for English proficiency assessments of spoken language. In this paper, we describe the development of an automatic pronunciation scorer built on state-of-the-art deep neural network models. The model is trained on a bespoke human-rated dataset that reflects current perspectives on pronunciation and intelligibility. The new scorer is evaluated along three criteria: How well it explains expert human ratings, how it compares to other state-of-the-art automatic pronunciation scorers in explaining expert human ratings, and the extent to which it exhibits bias toward different groups of test takers. Results indicate that the proposed scorer shows strong positive correlations with expert human ratings and outperforms other scorers. However, the scorer shows some bias related to audio quality and language family groups. We conclude with future directions for mitigating bias and argue that this scorer holds potential for use in operational settings.</p>","PeriodicalId":51371,"journal":{"name":"Language Learning","volume":"75 S1","pages":"170-203"},"PeriodicalIF":4.2,"publicationDate":"2025-07-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1111/lang.70000","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144629779","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"文学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：