From App to Fluent: A Bibliometric Analysis on the Impact of English Conversation Application on ESL Learners’ Skills

From App to Fluent: A Bibliometric Analysis on the Impact of English Conversation Application on ESL Learners’ Skills

Myka B. Patches¹, Joseline M. Santos², ^*Joyce Jenniefer T. Valeriano³

^1,3Graduate School, Bulacan State University, City of Malolos, Philippines

²College of Education, Bulacan State University, City of Malolos, Philippines

*Corresponding Author

DOI: https://dx.doi.org/10.47772/IJRISS.2025.903SEDU0251

Received: 28 April 2025; Accepted: 02 May 2025; Published: 03 June 2025

ABSTRACT

Purpose. This study examined the research landscape surrounding the use of English conversation applications (ECAs) and their impact on the speaking fluency of English as a Second Language (ESL) learners. It addresses the gaps in the existing literature by identifying influential studies, recurring themes, and emerging trends using bibliometric techniques.

Design/methodology/approach. A bibliometric analysis was conducted of 175 peer-reviewed articles sourced from the Scopus database (2015–2025). The data were analyzed using VOSviewer to visualize citation networks, co-authorship, keyword co-occurrence, and thematic clusters. The inclusion criteria focused on articles within the Social Sciences published in English. This study applied citation thresholds and thematic mapping to ensure analytical rigor.

Findings. The findings reveal steady growth in scholarly interest in ECAs, with key contributions highlighting their effectiveness in enhancing ESL learners’ fluency, pronunciation, confidence, and conversational competence. Research has been clustered around interactional linguistics, cognitive processing, corpus-based analysis, and digital communication. Core themes included turn-taking, prosody, discourse markers, and sociocultural influences on language use.

Practical implications. This study provides actionable insights for educators, curriculum designers, and app developers. This underscores the value of integrating ECAs into language instruction, tailoring them to diverse learner profiles, and supporting real-time feedback mechanisms. Policymakers and researchers should address underrepresented contexts and align technological tools with pedagogical goals.

Originality/value. This is the first comprehensive bibliometric analysis of ECAs in relation to ESL speaking fluency. This study offers a macro-level synthesis of research trends and contributes to the theoretical and practical understanding of how technology-mediated conversation supports fluency development. This study serves as a foundational resource for future inquiry and applications in language education.

Word count: 9,128 words, excluding references.

Funding Statement: This research received no specific grants from any funding agency in the public, commercial, or not-for-profit sectors.

Ethical Compliance: This study was a bibliometric analysis that utilized publicly available data from the Scopus database. As it did not involve human participants, personal data, or any form of intervention, ethical approval was not required. The research was conducted in accordance with the established ethical standards for non-interventional studies.

Data Access Statement: The bibliometric dataset analyzed in this study was obtained from the Scopus database. Data supporting these findings are available upon reasonable request from the corresponding author.

Conflict of Interest Declaration: The authors declare that they have no affiliations with or involvement in any organization or entity with any financial interest in the subject matter or materials discussed in this manuscript.

Author Contributions: The authors were solely responsible for the conceptualization, methodology, data collection, analysis, visualization, and writing of the manuscript.

Keywords: ESL learners, English conversation applications, fluency development, conversation analysis, mobile-assisted language learning, speaking skills

INTRODUCTION

Fluency in English speaking remains one of the most critical yet challenging goals for English as a Second Language (ESL) learners across the globe. While learners may excel in grammar, reading, and writing, many continue to struggle with oral fluency, marked by hesitation, limited vocabulary, unnatural intonation, and lack of spontaneity in real-time conversations (Nation, 2014). These difficulties are particularly pronounced in non-native English environments where authentic interaction opportunities are limited. In many ESL classrooms, speaking skills are underemphasized because of large class sizes, teacher-centered instruction, and exam-driven curricula that prioritize reading and writing over oral communication (Richards, 2008). Consequently, learners often graduate from ESL programs with insufficient fluency, hindering their ability to engage effectively in academic, professional, and social settings where English is required.

In response to these challenges, English conversation applications (ECAs) have emerged as innovative tools aimed at improving ESL learners’ speaking fluency. These applications, ranging from structured platforms like ELSA Speak and Babbel to socially interactive tools such as HelloTalk and Tandem, leverage mobile technology, artificial intelligence, and gamified learning environments to offer learners flexible, personalized, and immersive speaking practice. ECAs typically include features such as speech recognition, pronunciation feedback, simulated dialogue, and opportunities to converse with native or fluent speakers. The core idea behind these technologies is to supplement traditional classroom instruction with regular low-pressure conversational practice, which has long been recognized as essential for the development of fluency (Bygate, 2009).

Several studies have demonstrated the positive impact of ECAs on the speaking skills of ESL learners. For instance, Kim (2021) found that university students who used a mobile conversation app for eight weeks showed measurable improvements in their speech rate, pronunciation accuracy, and turn-taking ability. Similarly, Ahmadi (2018) argued that mobile-assisted language learning (MALL) applications significantly enhance oral proficiency, especially when learners are able to receive real-time feedback and repeat practice sessions at their own pace. Moreover, interactive ECAs foster learner motivation and autonomy, both of which are key components of effective language acquisition (Stockwell & Hubbard, 2013).

Furthermore, the use of ECAs in ESL contexts is linked to increased learner confidence and reduced speaking anxiety. In a study conducted by Wang and Chen (2020), learners who regularly used voice-based social apps reported feeling more comfortable initiating conversations and maintaining dialogues in English. The low-stakes nature of app-based communication combined with frequent exposure to spoken input and output provides a psychologically safe environment for learners to experiment with language. In addition, several ECAs employ speech recognition technology that allows learners to self-monitor their pronunciation and receive corrective feedback, thereby contributing to improved fluency and accuracy (Godwin-Jones, 2018).

Despite a growing body of research supporting the efficacy of ECAs, several gaps and limitations remain. First, most existing studies are context-bound, focusing on small, localized samples, often university students in East Asian countries, without accounting for broader demographic and geographic diversity (Kukulska-Hulme, 2020). This narrow scope limits the generalizability of the findings and fails to provide a comprehensive understanding of how ECAs function across different learner populations, age groups, and educational settings.

Second, methodological approaches in the literature tend to rely on pre- and post-intervention designs, learner surveys, or qualitative interviews. Although these methods provide valuable insights into learner perceptions and language gains, they do not offer a macro-level analysis of the field’s research trends, influential works, or thematic development over time. This limits the ability of educators, researchers, and developers to understand the evolution of scholarship in this area or identify areas that remain underexplored. For instance, little is known about the most frequently studied features of ECAs (e.g., pronunciation feedback, social interaction, gamification) or how different applications compare their pedagogical value and learner outcomes.

Third, few studies have attempted to map the intellectual landscape of research on ECAs and ESL-speaking fluency. There is a lack of bibliometric analyses that could illuminate the structure of this growing field by identifying core authors, journals, countries, institutions, and research networks. As the literature continues to expand, synthesizing and visualizing these scholarly contributions is essential for making sense of accumulated knowledge and guiding future inquiry (Donthu et al., 2021).

To address these gaps, this study conducted a comprehensive bibliometric analysis of existing research on the impact of English conversation applications on the fluency skills of ESL learners. Bibliometric analysis refers to the use of quantitative methods to examine patterns in scientific publications, allowing for a systematic and objective overview of the research landscape (Aria & Cuccurullo, 2017). This approach moves beyond narrative literature reviews by revealing not only what has been studied, but also how research in the field has developed over time, highlighting influential publications, recurring keywords, emerging themes, and collaborative relationships among researchers and institutions.

Specifically, the objectives of this study were to (1) identify the most prolific authors, journals, and institutions contributing to research on ECAs and ESL speaking fluency, (2) analyze keyword co-occurrence to determine core themes and research focuses, (3) visualize the co-authorship and citation networks that shape scholarly communication in this field, and (4) detect emerging trends and potential gaps that could inform future research directions. Using data retrieved from major academic databases such as Scopus and Web of Science, this study employed bibliometric tools such as VOSviewer and Bibliometric to produce visual maps and network diagrams that capture the intellectual structure of the literature.

Through this bibliometric approach, this study aims to contribute significantly to the understanding of how English conversation applications have been studied in relation to ESL learners’ fluency development. By offering a data-driven synthesis of the field, this research will provide educators, developers, and scholars with a clearer picture of where the field stands, where it is headed, and where further research is required. It also hopes to promote a more strategic and evidence-based integration of ECAs into ESL teaching practices, ultimately supporting learners in developing the fluency they need to succeed in an increasingly globalized world.

METHODOLOGY

This study employed a bibliometric analysis to investigate the academic landscape surrounding the use of English conversation applications and their impact on ESL learners’ fluency skills. The bibliometric approach provides a structured and quantitative means of identifying patterns, trends, and influential contributions within a specific body of the literature. For this analysis, VOSviewer, a widely recognized bibliometric mapping tool developed by Van Eck and Waltman (2010), was utilized to visualize co-authorship, citation networks, keyword co-occurrence, and thematic clusters. This tool allows for in-depth exploration of the intellectual and social structures of the research domain.

The bibliographic data for this study were sourced from the Scopus database, which is known for its comprehensive coverage of the peer-reviewed academic literature. A total of 175 documents were retrieved using the targeted search strategy. The search was limited to the years 2015 to 2025, capturing a decade of research that reflects the rise and evolution of mobile and digital conversation applications in the context of English language learning. The search terms used were “language,” “conversation,” and “communication,” which were selected to align with the focus of the study on spoken fluency, conversational interaction, and language acquisition through digital tools.

Several inclusion criteria were applied to ensure the quality and relevance of the dataset. First, only documents categorized under the Subject Area of Social Sciences were included because this field encompasses research in language education, linguistics, applied linguistics, communication studies, and educational technology. Second, only documents with the document type labeled as “Article” were selected, thereby excluding conference papers, book chapters, editorials, and other non-peer-reviewed materials. Third, the language of publication was restricted to English to maintain consistency in the content and analysis.

To refine the dataset for bibliometric mapping, the study applied specific thresholds to filter and focus the analysis on the most impactful and interconnected publications. The minimum number of citations per document was set to 12, ensuring that only articles with a notable degree of scholarly impact were included in the citation analysis. For co-citation analysis, the minimum number of citations of a cited reference was set at four, allowing the identification of foundational works frequently referenced across multiple studies. In the keyword co-occurrence analysis, a minimum occurrence threshold of 5 was used, meaning that only keywords that appeared in at least five documents were included in the visual mapping. This threshold was chosen to highlight commonly explored themes and concepts within the field, while filtering out isolated or marginal topics.

Once the data were cleaned and filtered, the selected articles were imported into VOSviewer for the analysis. The software generates network maps based on bibliographic coupling, co-authorship patterns, citation links, and keyword co-occurrence. Each map visualizes relationships between elements such as authors, sources, and terms, enabling the identification of key contributors, dominant themes, and emerging research trends. Clustering techniques within VOSviewer were employed to group the related items based on the strength of their bibliographic links. This facilitated the thematic categorization of the literature into meaningful clusters, revealing the underlying structure of research on English conversation applications in ESL education.

Through this methodological approach, this study aimed to provide a comprehensive and objective overview of scholarly activity on English conversation applications and their role in enhancing speaking fluency among ESL learners. The findings generated from this bibliometric analysis serve as a foundation for understanding not only what has been studied in this area, but also how the field has evolved over time, which scholars and institutions are leading the discourse, and where future research efforts may be directed.

Search Strategy and Data Collection

Table 1 Search string used for database search

RESULT AND ANALYSIS

This section presents the key findings derived from a bibliometric analysis of 175 documents related to conversation analysis and communication in English. The results highlight the growth and patterns of scholarly publications over time, identify the most-cited works that have significantly influenced the field, and uncover the intellectual structure of the research landscape through co-citation and co-word analysis.

Figure 1 Number of publications and citations

(Source: Scopus).

The Scopus trend analysis from 2015 to 2025 revealed that 175 documents on English conversation were published, reflecting a steady growth in scholarly interest in this area. In the initial years (2015–2016), the number of publications remained relatively low, with fewer than ten documents annually. However, in 2017, there was a notable increase, with approximately 18 documents published, signaling a rising academic focus on conversational dynamics, particularly in bilingual and multicultural contexts.

This upward trend experienced slight fluctuations in the following years. After a dip in 2018, the number of publications rose again in 2019, followed by a sharp decline in 2020, which saw the lowest output of the decade with approximately 11 documents. This drop may be attributed to global disruptions caused by the COVID-19 pandemic, which has temporarily impacted academic productivity across fields.

A strong rebound occurred in 2021, with over 20 publications, and the momentum continued to 2022 and 2023, peaking at approximately 25 documents. This surge corresponds with increased scholarly attention to language in digital and crisis-related communication, exemplified by highly cited works, such as Silesky (2023) and Seiter (2023), which explored language use during the pandemic.

The years 2024 and 2025 show a gradual decline in output, falling to 18 and 10 documents, respectively. This may reflect a shift in research focus or a typical delay in citation accumulation for new publications. Overall, the trend from 2015 to 2025 demonstrates a growing and sustained engagement with English conversation, especially in response to social, cultural, and global developments.

Citation analysis

By applying document citation analysis, Table 2 presents the highest-cited publications in the dataset. The top three most frequently cited works were Ribot and Krystal (2018) with 92 citations, Jones and Anna (2015) with 25 citations, and Kiramba and Lydiah Kananu (2023) with 22 citations. These citation counts indicate the influence and relevance of these studies in the academic community.

Table 2 Top 10 highest-cited documents

Trends and Emerging Themes in English Conversation

A bibliometric analysis of the top ten most-cited documents on English conversation reveals a diverse and evolving body of research that spans the sociocultural, cognitive, and technological dimensions of fluency development. Ribot (2018), the most cited work, highlighted how frequent and meaningful language use contributes significantly to expressive language growth in bilingual children, affirming the central role of authentic conversational engagement in fluency acquisition. Similarly, studies by Hammer (2017) and Kiramba (2023) explored the intersection of language, culture, and identity, especially among immigrant and bilingual student populations. These works emphasize that English conversation is deeply embedded in sociocultural contexts, influencing both linguistic performance and self-perception.

Another notable theme emerging from the analysis was the role of prosody and cognitive processing in fluent communication. Studies such as Johns (2021) and Tang (2021) have examined how prosodic features—intonation, rhythm, and stress—shape memory and processing in bilingual speech. Corps (2022) adds to this by challenging conventional views on turn-taking, suggesting that conversational gaps may offer limited insights into interactional fluency. These findings reflect growing interest in the real-time mechanisms that underlie spoken language fluency. The inclusion of Lind (2018), who studied lexical access in a bilingual speaker with dementia, extends this cognitive perspective to issues of language maintenance and decline across the lifespan.

Recent publications have highlighted the impact of digital communication environments on language use. For instance, Silesky (2023) and Seiter (2023) focused on social media interactions during the COVID-19 pandemic, addressing how misinformation, emotional expressions, and social support are navigated through online conversations. These studies signal an emerging trend in which English conversational fluency is analyzed not only in traditional settings, but also across virtual platforms, with implications for digital literacy and communication equity.

Overall, the top-cited literature reveals several key trends shaping the study of English conversation: a strong focus on bilingualism and identity, increased attention to prosody and cognitive processing, commitment to inclusive research involving diverse and vulnerable populations, and interest in the evolving nature of digital discourse. Collectively, these studies emphasize that English conversation is a multifaceted phenomenon shaped by context, interaction, and access. The findings suggest that future research and pedagogy should move toward models that are context-sensitive, inclusive, and aligned with real-world communication demands, ensuring that fluency is developed not just as a linguistic skill, but as a social and cognitive practice.

Citation Distribution and Implications for English Conversation

The citation trends in these studies suggest that research on English conversation is increasingly shaped by interdisciplinary perspectives that merge the linguistic, cognitive, sociocultural, and technological dimensions. The most highly cited study, Ribot (2018), with 92 citations, highlighted the pivotal role of consistent language use in developing expressive language skills among bilingual children, underscoring the long-standing emphasis on conversation as a key driver of fluency. Kiramba (2023) and Hammer (2017) explored the sociolinguistic aspects of fluency by examining how immigrant and bilingual individuals navigate identity and acculturation through language. These findings emphasize that conversational fluency is not only a linguistic achievement, but also a socially constructed process influenced by learners’ cultural and personal experiences.

The citation trends in these studies suggest that researchers also pay more attention to the cognitive and prosodic features of spoken interaction. For example, Johns (2021) and Tang (2021) investigated how prosody—intonation, rhythm, and stress—affects memory and processing load in bilingual speech, pointing to the importance of real-time processing in achieving fluent communication. Similarly, Corps (2022) challenges conventional metrics of conversational fluency, such as the duration of inter-speaker gaps, proposing that they may not reliably reflect turn-taking competence. This shift highlights the growing interest in how micro-level conversational cues contribute to perceived fluency.

In addition to these themes, studies such as Lind (2018) and Jones (2015) expanded the scope of English conversation research by addressing fluency within clinical and developmental contexts, including dementia and deafness, respectively. These contributions emphasize the need for inclusive communication models that consider diverse cognitive and sensory abilities. Meanwhile, Silesky (2023) and Seiter (2023) demonstrate a move toward understanding fluency and communication in digital and crisis contexts, such as misinformation during the COVID-19 pandemic and emotional expression on social media platforms.

The citation trends in these studies suggest that English conversation is increasingly viewed as a dynamic, multifaceted skill that is influenced by context, identity, interaction, and medium. The wide range of topics and disciplines represented in the top-cited documents indicate a field that not only expands in depth but also adapts to global, technological, and sociocultural changes. These findings imply that future research and pedagogy must continue to integrate authentic interactions, cognitive awareness, and inclusive strategies to support the development of English fluency across diverse learner populations and communicative environments.

Co-citation analysis

The top ten most co-cited documents in entrepreneurship and business education reveal the foundational literature shaping research in this interdisciplinary field, reflecting key themes such as entrepreneurial intention, pedagogy, innovation, and experiential learning. These highly co-cited works often emphasize the importance of entrepreneurial mindset development, particularly through active learning strategies and real-world engagement. Foundational studies, such as those by Krueger et al. and Ajzen, are frequently cited for their theoretical contributions to understanding entrepreneurial intention and planned behavior, forming a theoretical backbone for much of the research in this area.

Other frequently cited works focus on entrepreneurship education practices, highlighting the growing emphasis on pedagogical innovation such as project-based learning, design thinking, and the integration of social media into entrepreneurial instruction. Scholars, such as Rae and Gibb, underscore the value of embedding entrepreneurship within real-life contexts and experiential settings, promoting self-efficacy and creative problem-solving as essential learning outcomes.

In terms of methodology, co-citation patterns also indicate a balance between quantitative studies evaluating entrepreneurial outcomes and qualitative investigations into student perceptions, identity formation, and learning processes. The repeated appearance of works addressing sustainability and social entrepreneurship suggests a shift toward value-driven business education models that emphasize not only economic success, but also social impact.

Table 3 Top 10 documents with the highest co-citation and total link strength

Source: Author interpretation based on VOSviewer analysis

Based on network visualization, co-citation analysis produced four distinct clusters. figure 2 shows the network structure of the co-citation analysis. Each cluster was labelled and characterized based on representative publications according to the author’s inductive interpretation and understanding of the four clusters.

Figure 2 Co-citation analysis of the English Conversation and Fluency of ESL Learners

The co-citation analysis revealed four distinct clusters that highlight foundational influences in the study of English conversation. The red cluster, which appears most densely connected, centers on seminal works by Sacks, Schegloff, and Jefferson, who are widely regarded as pioneers of Conversation Analysis (CA). Key works such as “A Simplest Systematics for the Organization of Turn-Taking for Conversation” and Schegloff’s work on turn organization appear as central nodes, indicating their frequent co-citations and lasting impacts. Also included in this cluster are Sidnell (Conversation Analysis), Heritage, and Garfinkel, whose ethnomethodological perspectives reinforce the foundational CA tradition.

The green cluster continues this thread, focusing on practical extensions of CA, including Jefferson’s “Glossary of Transcript Symbols” and further studies on turn-taking, underscoring the technical and methodological relevance of these works for empirical research.

The yellow cluster highlights corpus-based approaches, with the International Corpus of English (ICE) serving as a key node. This suggests a methodological shift in some areas of the field toward large-scale data analysis, bridging qualitative and quantitative traditions.

The blue cluster, though smaller, features Clark’s (1996) “Using Language”, representing psycholinguistic and cognitive-pragmatic perspectives. Its linkage to other clusters suggests an interdisciplinary connection between interactional linguistics and cognitive sciences.

Overall, this co-citation map reflects how English conversation research is rooted in classic CA theory while simultaneously expanding into corpus linguistics and cognitive approaches, demonstrating the field’s evolving and integrative nature.

• Cluster 1 (red): The red cluster in the co-citation analysis represents the theoretical core of Conversation Analysis (CA), a field grounded in ethnomethodology and concerned with the structure of everyday spoken interaction. At its center are seminal works by Sacks, Schegloff, and Jefferson, whose influential study on turn-taking (A Simplest Systematics for the Organization of Turn-Taking for Conversation, Sacks, Schegloff, & Jefferson, 1974) established fundamental principles for understanding how speakers coordinate turns in conversation. These authors emphasized the orderly nature of talk-in-interaction, showing that even casual dialogue follows systematic patterns. Also featured in this cluster is Jefferson’s (2004) contribution to transcription conventions, which standardized the representation of spoken features such as pauses, overlaps, and intonation—crucial for rigorous CA research. Heritage (1984) further advanced CA by exploring institutional talk and the organization of social actions, while Sidnell (2010) provided a comprehensive overview of CA’s analytical framework and its relevance across diverse sociolinguistic contexts. Garfinkel’s (1967) ethnomethodological influence is also evident as CA builds upon his view that social order is produced through participants’ methods of sense-making. Collectively, this cluster underscores the discipline’s focus on micro-level interactions, examining how meaning, identity, and social structure are negotiated. It remains a cornerstone in English conversation research, providing a conceptual and analytical foundation upon which other approaches such as corpus linguistics and cognitive pragmatics are later built. The prominence of this cluster in co-citation patterns reflects the enduring influence of the CA’s qualitative and theory-driven orientation.
• Cluster 2 (green): The green cluster in the co-citation analysis builds upon the foundational theories of Conversation Analysis (CA) by focusing on methodological development and analytical refinement. At the center of this cluster is Gail Jefferson’s (2004) widely used Glossary of Transcript Symbols, which standardizes the transcription of conversational features, such as pauses, overlapping speech, laughter, and intonation. This transcription system has become essential for researchers conducting fine-grained empirical analyses of talk-in-interaction Central to this cluster is Schegloff (2007) work on sequence and turn organization, which offers a deeper theoretical account of how conversational turns are structured and how participants coordinate conversations across extended sequences. These methodological contributions enable a more systematic and replicable approach to analyzing spoken discourse, particularly in institutional and cross-cultural contexts.

The cluster also reflects a broader concern with applied conversation analysis, showing how foundational CA concepts are operationalized in diverse research settings, from clinical and educational interactions to media and digital communication. Researchers, such as Sidnell (2010) and Heritage (2005), continue to bridge theory and application by demonstrating how these tools illuminate subtle aspects of meaning-making and social action. The green cluster represents the practical arm of CA, equipping scholars with the tools needed for consistent data interpretation. Its high level of co-citation with core CA texts highlights the integration of methodological rigor with theoretical depth, reinforcing CA’s position as both a theoretical framework and research methodology in English conversation studies.

• Cluster 3 (blue): The blue cluster in the co-citation analysis highlights the cognitive-pragmatic and psycholinguistic dimensions of English conversation research. Central to this cluster is Clark’s (1996) influential work Using Language, which emphasizes how communication relies on shared knowledge, joint attention, and management of common ground between speakers. Unlike Conversation Analysis, which focuses on the structure of interaction from an external viewpoint, this cluster is concerned with internal cognitive processes that underlie conversational behavior, such as inference, memory, attention, and intention recognition.

This cluster aligns with pragmatic theories of meaning and interaction, including Grice’s (1975) cooperative principle and theories of relevance (Sperber & Wilson, 1995), although it tends to frame these within cognitive communication models. Studies of this tradition often investigate how speakers produce and comprehend utterances in real time, manage ambiguity, and interpret implied meaning—issues central to both language processing and social cognition. This perspective is particularly relevant in exploring how conversation functions in atypical or high-stakes contexts such as second-language acquisition, cognitive impairment, or digital interaction.

Although smaller in size, the blue cluster plays a crucial interdisciplinary role by bridging linguistics, psychology, and communication studies. Its connections to other clusters suggest that scholars increasingly recognize the importance of integrating mental representations and speaker intentions into broader analyses of spoken discourse. Overall, the blue cluster adds an essential cognitive layer to English conversation research, enriching our understanding of how interactions are mentally constructed, interpreted, and coordinated.

• Cluster 4 (yellow): The yellow cluster in the co-citation analysis reflects a shift in English conversation research toward corpus-based and quantitative approaches. At its core is the International Corpus of English (ICE) project, initiated by Nelson, Wallis, and Aarts (2002), which provides structured, annotated data from varieties of English around the world. Unlike traditional Conversation Analysis (CA), which relies on a detailed examination of limited interactional data, corpus linguistics enables the analysis of large-scale patterns in spoken language. This cluster emphasizes frequency, collocation, discourse markers, and pragmatic features within naturally occurring speech, contributing to empirical generalizations of English usage across contexts and regions.

The yellow cluster often draws on spoken corpus data to explore conversational routines, discourse strategies, and grammatical variations. Researchers in this area such as Carter and McCarthy (2006) focus on identifying lexical bundles, backchannels, and formulaic expressions that structure everyday talk. This approach aligns more closely with functional linguistics and computational methods, offering a complementary perspective to CA’s microanalytic lens. Importantly, this cluster broadens the scope of English conversation research by incorporating cross-cultural, intercultural, and World English perspectives, illustrating how conversational norms shift across sociolinguistic environments.

Although methodologically distinct from foundational CA, the yellow cluster is not oppositional. Instead, it offers tools for scaling up interactional analysis, allowing researchers to identify macro patterns that CA alone may overlook. Its co-citation presence alongside CA texts signals the convergence of qualitative and quantitative paradigms, enriching the field with methodological diversity and global applicability.

Table 4 summarizes the co-citation analysis by presenting its clusters, cluster labels, number of articles, and representative publications.

Table 4 Co-citation clusters on the role of Educational tools in business education

Cluster 1 explores how Conversation Analysis (CA) and ethno-methodology provide foundational frameworks for understanding the development of English fluency through interaction. Drawing from ethnomethodology, Heritage (1984) emphasized the idea that everyday conversation is not chaotic but methodically organized, providing learners with structured environments to practice and internalize interactional norms. This notion underpins CA’s relevance in second language contexts, where fluency is not merely linguistic accuracy, but the ability to participate in orderly, context-sensitive interactions.

The seminal work by Sacks, Schegloff, and Jefferson (1974) on turn-taking systems reveals that fluency involves mastering the timing and organization of speech—skills that are vital for engaging in spontaneous English conversation. Their identification of rules for turn allocation and transition relevance places fluency within a framework of interactive competence and not just lexical knowledge.

Schegloff (2007) expands on this by detailing how sequences in interactions (e.g., adjacency pairs, repairs) function as building blocks of conversation. This sequential organization is essential for learners to develop interactional fluency as it shapes their ability to anticipate responses, manage breakdowns, and sustain dialogue.

Sidnell (2010) synthesized these ideas in a comprehensive introduction to CA, making the field accessible to applied linguists and educators. His work helps bridge theory and practice by demonstrating how analyzing authentic interactions can inform fluency-focused pedagogies.

Together, these publications form a theoretical cluster that reframes fluency as a socially organized, interactional achievement developed through participation in structured conversational practices (Heritage, 1984; Sacks et al., 1974; Schegloff, 2007; Sidnell, 2010).

Cluster 2 centers on the structural foundations of conversation, emphasizing how turn-taking mechanisms and transcription conventions inform our understanding of spoken fluency. The seminal work of Sacks, Schegloff, and Jefferson (1974) introduced the systematic organization of turn-taking in conversation, identifying the rules that govern speaker transitions. For language learners, fluency is not only about linguistic accuracy, but also about timely and appropriate turn-taking, a critical skill for conversational competence.

Schegloff (1996) deepened this understanding by linking grammar to interactional practices, showing how syntactic structures are shaped by and contribute to the flow of conversation. This intersection suggests that fluency includes not only sentence formation, but also the ability to embed grammar within socially meaningful interactions.

Jefferson (2004) complements these theoretical insights by offering a detailed transcription system that captures nuances of talk, such as pauses, overlaps, and intonation, which are essential indicators of fluency. Accurate transcription enables researchers and educators to analyze the rhythm and coordination of spoken interactions in detail.

Together, these works provide a methodological and analytical toolkit for studying how learners acquire fluency through the mastery of conversational structures, making turn-taking and transcription central to both research and pedagogy in spoken English development.

Cluster 3 highlights the essential role of real-time conversational dynamics in developing students’ fluency in English. Clark (1996) emphasized that language use in conversation is a collaborative, context-driven activity in which meaning is jointly constructed through turn-taking and mutual understanding. His theory suggests that fluency stems not only from lexical or grammatical knowledge but also from students’ ability to engage in spontaneous, coordinated dialogue. Levinson and Torreira (2015) deepened this perspective by analyzing the cognitive demands of turn-taking, showing that conversational fluency requires rapid language processing, often within milliseconds. This real-time coordination challenges learners in planning speech, anticipating responses, and managing timing effectively. Together, these studies argue that English conversation fosters fluency by pushing learners to operate under real-time constraints, developing both linguistic competence and processing efficiency. The cluster underscores that fluency is not only about speed but also responsiveness and coherence in natural spoken interaction.

Cluster 4 emphasized the role of discourse organization in fostering student fluency in English conversations. Schiffrin (1987) explored the function of discourse markers, such as well, so, and you know, as crucial elements that structure spoken interaction, aiding in coherence, topic management, and speaker-listener coordination. Her work suggests that fluency involves more than speed or accuracy; it depends on learners’ ability to use these markers to navigate and negotiate meaning in real-time conversations. The International Corpus of English (ICE) complements this by providing authentic spoken data across global English varieties, highlighting how discourse structures vary and how fluency manifests in different contexts. Together, these studies suggest that fluency is deeply tied to students’ command of discourse-level features, not just sentence-level grammar. This cluster demonstrates that developing fluency involves learning how to structure conversation coherently and contextually using discourse markers and patterns drawn from real-world language use.

Co-word analysis

Co-word analysis was applied to the same database. From the 1,620 keywords, 49 met the minimum of six occurrences, resulting in two clusters. The keywords with the highest co-occurrence were student (177), human (159), and humans (125). Table 5 summarizes the top 15 co-occurring keywords with their number of occurrences and total link strength.

Table 5 Top 15 keywords in the co-occurrence of keywords analysis

Co-occurrence keyword analysis highlights the multidimensional nature of research on English conversation and fluency, integrating linguistic, cognitive, demographic, and methodological themes. The most prominent keyword, conversation (111 occurrences; total link strength: 639), underscores its central role in fluency development, reinforcing the idea that authentic spoken interaction is key to mastering fluent communication (Sacks et al., 1974). Closely tied to this are language (67; 540), speech (32; 362), and interpersonal communication (29; 320), all of which suggest that fluency is not merely linguistic competence but the ability to manage real-time, socially embedded exchanges (Schegloff, 1996; Clark, 1996).

The presence of human (88; 868), humans (55; 613), female (51; 606), male (47; 557), and adult (37; 439) indicates that participant demographics were frequently considered, suggesting that fluency development may vary by age and gender. Similarly, the inclusion of child (24, 282) points to developmental perspectives in language acquisition research.

Methodologically, terms such as article (55, 599), controlled study (20, 259), clinical article (17, 215), and human experiment (29, 310) reflect a strong empirical orientation, with researchers using experimental and clinical designs to investigate fluency outcomes. The keyword communication (48, 310) broadly links these strands, suggesting that fluency is increasingly seen as a multimodal, interactive process.

Overall, this keyword cluster reveals a dynamic research field in which conversation, human interaction, and empirical analysis intersect to explain the development and study of English fluency.

Figure 3 presents a network map of the co-word analysis. The map produced four clusters that were classified and labeled based on the author’s inductive interpretation of the occurring words. All clusters were closely related and partially integrated.

Figure 3 Co-word analysis on the Role of Educational Tools in Business Education

• Cluster 1 (red): This cluster centers on the keywords: conversation, speech, interpersonal communication, and language, highlighting how active engagement in spoken interaction contributes to fluency. The prominence of conversation (111 occurrences) emphasizes its foundational role in building fluency through real-time dynamic exchanges. As Sacks et al. (1974) and Clark (1996) argued, fluency emerges through structured yet spontaneous dialogue, where learners practice turn-taking, negotiation of meaning, and rapid language processing. The inclusion of interpersonal communication and speech suggests a focus on communicative competence and framing fluency as both linguistic and social skills refined through repeated, authentic interaction.
• Cluster 2 (green): Keywords such as human (88), humans (55), human experiment (29), and adult (37) indicate a strong cognitive and psychological orientation. This cluster reflects research on how human cognitive processes such as memory, attention, and processing speed affect the acquisition of fluency in conversation. Levinson and Torreira (2015) provide relevant insight, demonstrating that conversational timing requires rapid planning and real-time comprehension. The frequent mention of human experiment and adult suggests that experimental designs are often used to measure fluency performance and to investigate how adults, in particular, manage conversational demands.
• Cluster 3 (blue): This cluster reflects a demographic focus with keywords such as female (51), male (47), child (24), and adult (37). The recurrence of these terms suggests that fluency development should be examined across age and gender lines. Research within this cluster often explores how gendered interactional styles or developmental stages (e.g., childhood vs. adulthood) influence fluency. For example, studies may investigate whether females engage in more supportive conversational strategies that promote fluency or how children’s fluency acquisition patterns differ from those of adults in structured versus unstructured settings (Pavlenko, 2004).
• Cluster 4 (yellow): Keywords such as article (55), controlled study (20), clinical article (17), and human experiment (29) indicate a strong methodological backbone. This cluster is characterized by its reliance on quantitative, experimental, and clinical research designs to study fluency. Controlled studies often manipulate conversational variables such as task type or feedback to observe fluency outcomes. This aligns with task-based learning research and fluency assessments that aim to quantify speech rate, pause length, or syntactic complexity in real-time English conversations (Skehan, 2009).
• Cluster 5 (violet): The presence of communication (48), linked to both linguistic and social terms, frames fluency as a multimodal interactive skill. This cluster integrates diverse elements from conversation to discourse markers and gestures, suggesting that effective fluency involves more than verbal accuracy. Schiffrin (1987) emphasized the role of discourse markers in structuring interactions, while Gumperz (1982) highlighted how contextual cues shape conversational meaning. This cluster views fluency as deeply embedded in the social context and shaped by how learners manage spoken language across different communicative modes.

Table 6 summarizes the co-word analysis represented by the cluster label, number of keywords, and representative keywords.

Table 6: Co-word analysis on technology tools in business education

Implications of the Study

The findings of this bibliometric analysis offer important insights into both the theoretical and practical relevance of the digital tools such as English conversation applications in education. As educational institutions increasingly integrate digital tools and pedagogical innovations into their curricula, understanding the trends, frameworks, and applications of these tools becomes essential. This study contributes to ongoing discussions on how technology-enhanced learning can support student engagement. The following section outlines the key theoretical and practical implications of this study.

Theoretical Implications

This bibliometric analysis offers key theoretical insights into how English fluency develops among ESL learners, especially through conversation applications. A central implication is the reframing of fluency as an interactional and social competence rather than merely a linguistic or grammatical skill. Drawing from Conversation Analysis (Sacks, Schegloff, & Jefferson, 1974), fluency involves effectively navigating structured yet spontaneous interactions through turn-taking, sequence organization, and repair mechanisms (Schegloff, 2007; Sidnell, 2010).

This view is reinforced by sociocultural theory, particularly Vygotsky’s (1978) Zone of Proximal Development, which highlights the role of social interaction in learning. Conversation apps act as mediating tools that provide scaffolded, meaningful communication—via native speakers, AI, or contextual cues—pushing learners slightly beyond their current abilities (Lantolf & Thorne, 2006). These platforms align with theories viewing language acquisition as socially mediated and context-sensitive.

Cognitive-pragmatic theories further support this multidimensional model of fluency. Clark (1996) and Levinson & Torreira (2015) underscore the importance of joint attention, shared meaning, and the cognitive demands of rapid turn-taking. Apps that simulate real-time interaction help learners develop the processing speed and coordination needed for fluent conversation.

Corpus linguistics and discourse-based models also contribute, shifting focus to discourse markers and interactional features—such as fillers, hedges, and turn initiators—as key components of strategic competence (Schiffrin, 1987; Canale & Swain, 1980). These elements enhance conversational coherence and fluidity beyond grammar and vocabulary alone.

Additionally, the rise of digital and multimodal communication reshapes fluency to include asynchronous, multimedia interaction. Studies (Silesky, 2023; Seiter, 2023) emphasize that fluency now encompasses digital literacy and the ability to interpret voice, text, emojis, and contextual signals—aligning with calls to expand communicative competence frameworks (Blake, 2013; Thorne, 2013).

Demographic variables—such as age, gender, and cognitive profiles—also emerged as significant, highlighting the need for inclusive fluency models that account for learner diversity (Lind, 2018; Jones, 2015; Pavlenko, 2004). This reflects constructivist perspectives that emphasize individual differences in learning trajectories.

Finally, the prevalence of controlled, clinical, and corpus-based studies reflects a shift toward empirically grounded theory-building in applied linguistics. These methods support a data-informed understanding of fluency acquisition and reinforce the importance of iterative theory development (Donthu et al., 2021; Aria & Cuccurullo, 2017).

In sum, this study shows that English fluency is a multidimensional construct shaped by interactional, cognitive, sociocultural, and digital factors. Conversation apps not only aid practical skill development but also help redefine fluency theories to better reflect the realities of global and technologically mediated communication.

Practical Implications

The results of this bibliometric analysis offer practical implications for educators, curriculum designers, app developers, and policymakers seeking to improve ESL learners’ speaking fluency through technology-enhanced learning. One of the key findings reveals that English conversation applications (ECAs) serve not only as supplementary tools but also as essential platforms for enabling authentic, interactive language practice in low-stakes environments. As highlighted by Bygate (2009) and Kim (2021), regular exposure to real-time conversational scenarios using apps, such as HelloTalk or ELSA Speak, enhances learners’ speech rate, pronunciation, and turn-taking competence. Educators can leverage these findings by integrating ECAs into classroom instruction or blended learning models, particularly to address the lack of real-world speaking opportunities in traditional ESL settings (Richards, 2008).

Furthermore, the analysis underscores the need for pedagogical designs that recognize the cognitive and interactional dimensions of fluency. Citing works by Clark (1996) and Levinson and Torreira (2015) show that effective fluency development depends on learners’ ability to process language in real time, respond appropriately in conversations, and manage prosodic and timing elements. This suggests that language programs should prioritize interactive speaking tasks and AI-supported feedback systems that mirror the immediacy and responsiveness of natural conversations. App developers are encouraged to incorporate features, such as adaptive speech feedback, conversation branching, and discourse-level prompts, to simulate the timing and rhythm of real interactions, which are critical for building both fluency and confidence (Godwin-Jones, 2018).

Additionally, the analysis revealed that fluency is not one-size-fits-all, with demographic factors such as age and gender playing roles in how learners develop conversational skills. Given the presence of terms, such as child, adolescent, adult, and gender in the keyword co-occurrence map, it is clear that learning needs differ across user groups. Practitioners should, therefore, adapt ECA-based activities based on learner profiles, offering differentiated pathways for younger learners, adults, or learners with limited exposure to English (Lind, 2018). For instance, apps targeting children may need to include gamified elements and scaffolded dialogue, while tools for adult learners might benefit from scenario-based interactions rooted in the workplace or academic communication.

These findings also highlight the importance of inclusive and socially relevant content. Highly cited studies, such as those by Kiramba (2023) and Hammer (2017), point to the intersection of language use, identity, and sociocultural experience, especially for immigrant and bilingual populations. Practically, this means that ECAs and instructional materials should reflect culturally diverse narratives and provide learners with the opportunity to express their identities through language. Moreover, language educators are encouraged to facilitate discussions that connect learners’ personal experiences with the language they are acquiring, making fluency development not just linguistic but also empowering.

Another major implication is the need to incorporate digital literacy into ESL instruction. The increased attention paid to language use in virtual environments, as seen in studies by Silesky (2023) and Seiter (2023), demonstrates that fluency today extends beyond face-to-face communication. ESL learners must develop the ability to navigate digital communication platforms from video calls to social networking sites where conversational norms differ. This suggests that fluency-focused curricula should include training in online interactions, netiquette, and hybrid discourse styles. Teachers should model and practice digital speaking tasks, such as participating in virtual discussions or simulating chatbot dialogues, to prepare students for communication demands in the 21st century.

Lastly, the use of bibliometric tools, such as VOSviewer, to map research trends has practical implications for research planning and policy development. The ability to identify high-impact studies, leading institutions, and emerging topics provides a strategic overview of stakeholders aiming to fund or develop language-learning programs. Policymakers can use such data to allocate resources to underexplored areas, such as conversational fluency in marginalized populations, or to support the development of open-access ECAs aligned with the best practices identified in the literature (Donthu et al., 2021; Aria & Cuccurullo, 2017). Future research should also consider collaborations between app developers and educators to ensure that the pedagogical design of ECAs is rooted in both theoretical and real-world classroom needs.

In summary, bibliometric analysis offers compelling evidence that English conversation applications are a valuable, evidence-based solution for improving ESL learners’ fluency. By integrating these tools into teaching practice, designing learner-centered app features, and aligning content with cognitive and sociocultural realities, stakeholders in language education can significantly enhance learners’ communicative competence in an increasingly digital and interconnected world.

CONCLUSION, LIMITATIONS, AND FUTURE RECOMMENDATIONS

Several key conclusions can be drawn based on the findings of the bibliometric analysis. This study highlights a growing scholarly interest in the integration of English conversation applications within ESL learning environments. This surge in academic attention mirrors the increasing global reliance on digital tools to enhance language acquisition, particularly in speaking and listening. Evidence from the reviewed literature demonstrates that conversation-based applications positively impact ESL learners’ fluency, pronunciation, and confidence, enabling them to communicate effectively in real-world situations. The analysis also revealed geographical patterns, with significant contributions coming from regions where English education is a national priority, such as Southeast Asia and the Middle East. Moreover, applications such as Duolingo, HelloTalk, and Tandem are frequently referenced, suggesting their strong influence on ESL teaching practices owing to their interactive nature and real-time feedback capabilities.

Despite these insights, this study has some limitations. The scope of the analysis was confined to specific databases and search terms, which may have led to the exclusion of relevant literature published under different terminologies or non-indexed sources. Additionally, reliance on citation metrics and publication counts means that the analysis may overlook the qualitative depth and educational impact of individual studies. There is also a language bias, as the majority of reviewed studies were written in English, potentially neglecting valuable research conducted in other languages. Furthermore, given the rapid pace of technological advancement, the relevance of the findings may diminish quickly as newer applications and tools are developed.

To address these limitations and build on the current findings, several recommendations for future research are proposed. Scholars should consider integrating bibliometric analyses with qualitative approaches to provide a more comprehensive understanding of how English conversation applications influence language learning. Expanding the geographical and cultural scope of these studies would offer richer and more diverse perspectives on the effectiveness of these tools. Longitudinal research is encouraged to determine the sustained impact of app-based learning on language proficiency over time. In addition, future investigations should delve into which specific app features, such as gamification, artificial intelligence-driven feedback, or peer-to-peer interaction, are most beneficial for learners. Finally, exploring the role of educators in guiding the use of these applications could yield valuable insights into how formal instruction and digital tools work synergistically to support ESL learners.

REFERENCES

1. Ahmadi, M. R. (2018). The use of technology in English language learning: A literature review. International Journal of Research in English Education, 3(2), 115–125. https://doi.org/10.29252/ijree.3.2.115
2. Aria, M., & Cuccurullo, C. (2017). Bibliometrix: An R-tool for comprehensive science mapping analysis. Journal of Informetrics, 11(4), 959–975. https://doi.org/10.1016/j.joi.2017.08.007
3. Bygate, M. (2009). Teaching and testing speaking. In M. Long & C. Doughty (Eds.), The handbook of language teaching (pp. 412–440). Wiley-Blackwell.
4. Clark, H. H. (1996). Using language. Cambridge University Press. https://doi.org/10.1017/CBO9780511620539
5. Donthu, N., Kumar, S., Mukherjee, D., Pandey, N., & Lim, W. M. (2021). How to conduct a bibliometric analysis: An overview and guidelines. Journal of Business Research, 133, 285–296. https://doi.org/10.1016/j.jbusres.2021.04.070
6. Godwin-Jones, R. (2018). Using mobile technology to develop language skills and cultural understanding. Language Learning & Technology, 22(3), 1–17. https://www.lltjournal.org/item/3133
7. Hammer, K. (2017). They speak what language to whom?!: Acculturation and language use for communicative domains in bilinguals. Bilingual Research Journal, 40(1), 40–55. https://doi.org/10.1080/15235882.2016.1276028
8. Heritage, J. (1984). Garfinkel and ethnomethodology. Polity Press. https://politybooks.com/bookdetail/?isbn=9780745600056
9. Jefferson, G. (2004). Glossary of transcript symbols with an introduction. In G. H. Lerner (Ed.), Conversation analysis: Studies from the first generation (pp. 13–31). John Benjamins.https://doi.org/10.1075/pbns.125.02jef
10. Jones, A. C. (2015). Confronting the language barrier: Theory of mind in deaf children. Journal of Deaf Studies and Deaf Education, 20(3), 268–277. https://doi.org/10.1093/deafed/env009
11. Kim, H. (2021). Enhancing ESL learners’ speaking fluency through mobile-assisted conversation practice. Computer Assisted Language Learning, 34(6), 1–23. https://doi.org/10.34190/ejel.21.3.2974
12. Kiramba, L. K. (2023). “It’s OK. She Doesn’t Even Speak English”: Narratives of language, culture, and identity negotiation by immigrant high school students. Journal of Language, Identity & Education, 22(1), 1–15. https://doi.org/10.1177/0042085919873696
13. Kukulska-Hulme, A. (2020). Mobile-assisted language learning [MALL]. In M. R. Apple & D. Da Silva (Eds.), The Routledge handbook of language and digital communication (pp. 234–248). Routledge. https://oro.open.ac.uk/57023/1/ userdata documents5 ak35 Desktop Accepted%20 Manuscript Concise%20Encyclopedia.pdf
14. Lind, M. (2018). Lexical access in a bilingual speaker with dementia: Changes over time. Aphasiology, 32(8), 891–910. https://pubmed.ncbi.nlm.nih.gov/29043848/
15. Nation, I. S. P. (2014). Learning vocabulary in another language (2nd ed.). Cambridge University Press. https://doi.org/10.1017/CBO9781139858656
16. Richards, J. C. (2008). Teaching listening and speaking: From theory to practice. Cambridge University Press. https://www.professorjackrichards.com/wp-content/uploads/teaching-listening-and-speaking-from-theory-to-practice.pdf
17. Ribot, K. M. (2018). Language use contributes to expressive language growth: Evidence from bilingual children. Child Development, 89(2), 523–537. https://doi.org/10.1111/cdev.12753
18. Sacks, H., Schegloff, E. A., & Jefferson, G. (1974). A simplest systematics for the organization of turn-taking for conversation. Language, 50(4), 696–735. https://doi.org/10.1353/lan.1974.0010
19. Schegloff, E. A. (2007). Sequence organization in interaction: A primer in conversation analysis I. Cambridge University Press. https://psycnet.apa.org/record/2006-13301-000
20. Schegloff, E. A. (1996). Turn organization: One intersection of grammar and interaction. In E. Ochs, E. A. Schegloff, & S. A. Thompson (Eds.), Interaction and grammar (pp. 52–133). Cambridge University Press. https://www.cambridge.org/core/books/abs/interaction-and-grammar/turn-organization-one-intersection-of-grammar-and-interaction/972375B8AB5D221FDA5EEEFD52AA9239
21. Schiffrin, D. (1987). Discourse markers. Cambridge University Press. https://doi.org/10.1017/CBO9780511611841
22. Seiter, C. R. (2023). Social support and aggressive communication on social networking sites during the COVID-19 pandemic. Journal of Computer-Mediated Communication, 28(1), 134–150. https://www.researchgate.net/publication/349461219_Social_Support_and_Aggressive_Communication_on_Social_Network_Sites_during_the_COVID-19_Pandemic
23. Sidnell, J. (2010). Conversation analysis: An introduction. Wiley-Blackwell. https://www.wiley.com/en-us/Conversation+Analysis%3A+An+Introduction-p-9781405159005
24. Silesky, M. D. (2023). A multifaceted campaign to combat COVID-19 misinformation in the Hispanic community. Health Communication, 38(3), 217–228. https://pmc.ncbi.nlm.nih.gov/articles/PMC9673890/
25. Stockwell, G., & Hubbard, P. (2013). Some emerging principles for mobile-assisted language learning. Monterey, CA: The International Research Foundation for English Language Education. https://www.tirfonline.org/wp-content/uploads/2013/11/TIRF_MALL_Papers_StockwellHubbard.pdf
26. Tang, K. (2021). Prosody leaks into the memories of words. Journal of Memory and Language, 118, 104220. https://pubmed.ncbi.nlm.nih.gov/33508575/
27. Van Eck, N. J., & Waltman, L. (2010). Software survey: VOSviewer, a computer program for bibliometric mapping. Scientometrics, 84, 523–538. https://doi.org/10.1007/s11192-009-0146-3
28. Vygotsky, L. S. (1978). Mind in society: The development of higher psychological processes. Harvard University Press. https://www.jstor.org/stable/j.ctvjf9vz4
29. Wang, Y., & Chen, N.-S. (2020). Effects of conversational strategies on English speaking anxiety in virtual reality–based settings. Educational Technology Research and Development, 68, 1971–1989.