CoChat
Sign Up
Cited by 143Speech and dialogue systemsIntelligent Tutoring Systems and Adaptive LearningEFL/ESL Teaching and Learning

A systematic review of research on speech-recognition chatbots for language learning: Implications for future directions in the era of large language models

Jeon, Jaeho, Lee, Seongyong, Choi, Seongyune·2024·Interactive Learning Environments
DOI 10.1080/10494820.2023.2204343
Open this paper in CoChat — free account

Abstract

Chatbot research has received growing attention due to the rapid diversification of chatbot technology, as demonstrated by the emergence of large language models (LLMs) and their integration with automatic speech recognition. However, among various chatbot types, speech-recognition chatbots have received limited attention in relevant research reviews, despite their increasing potential for language learning. To fill this gap, 32 empirical studies on speech-recognition chatbots for language learning were reviewed. The following information was reviewed for each study: basic publication information, research focus, location of chatbot use, methodology, group design format, participant information, intervention duration, target language, device type adopted, and chatbot role. An upward trend in research quantity starting in 2020 was identified, which accelerated exponentially in 2022. College students were more likely than other groups to be involved in research, and English as a second or foreign language was the most common target language. Most studies focused on participants’ perceptions of chatbots and the degree to which using chatbots helped them develop their speaking or listening proficiency. Methodologically, single-chatbot design using mixed methods was the most common design format, and most studies were conducted for more than one month in laboratory or classroom settings. Conventional mobile devices, such as smartphones, tablet PCs, and smart speakers without a screen, were the most frequently adopted device types. The chatbots’ most common role was as conversational partner. A detailed discussion of these results and their implications for future research on speech-recognition chatbots, particularly regarding the use of LLM-powered chatbots, is provided.

Integrity signals

Registry

Crossref · 10.1080/10494820.2023.2204343

Cites

13 of 13 references verified0 citation errors · 0 not checkable

Statistics

Recomputing p-values needs the paper’s full text.

Claims the paper makes

This paper’s full text isn’t openly available, so its claims and self-check weren’t run. The registry and citation checks still apply.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • Reviews targeting language education typically focus on potential benefits, affordances, challenges, future directions, attitudes, and overview of current state.

    “Reviews targeting language education typically focus on the potential benefits, affordances, challenges, future directions for AI applications, attitudes toward AI, and an overview of the current state of research.”

    From Generative AI in Language Education: Bridging Divide and Fostering Inclusivity (2025) · cites this paper as Jeon et al., 2023

  • Reviews targeting language education do not concentrate exclusively on how GenAI can be leveraged to narrow educational gaps and promote inclusivity in language education.

    “However, they do not concentrate exclusively on how GenAI can be leveraged to narrow educational gaps and promote inclusivity in language education-an area that warrants further exploration.”

    From Generative AI in Language Education: Bridging Divide and Fostering Inclusivity (2025) · cites this paper as Jeon et al., 2023

  • Using ChatGPT for English learning has big potential because learners admitted it can expand vocabulary, enhance grammatical and syntactical structures, and improve written and conversational skills.

    “In addition, as the results of this study show, using ChatGPT for English learning has a big potential since learners admitted that ChatGPT can expand their vocabulary, enhance their grammatical and syntactical structures, and thus improve their written and conversational skills [28,29,41].”

    From Assessing the Usability of ChatGPT for Formal English Language Learning (2023) · cites this paper as [28]

  • Students can ask questions, share thoughts, and discuss various topics, which reflects real-life dialogues and makes them motivated to learn a foreign language.

    “Students can ask questions, share their thoughts, and discuss various topics, which reflects real-life dialogues and makes them motivated to learn a foreign language (cf. [28,29]).”

    From Assessing the Usability of ChatGPT for Formal English Language Learning (2023) · cites this paper as [28]

Cites

Flagged first, then by citations

Cited by

Sorted by citations

Authors
Jeon, J., Lee, S., & Choi, S.
Affiliations
  • Indiana University Bloomington (US)
  • Hannam University (KR)
  • Korea University (KR)
Corresponding author
Lee, Seongyong
ORCID
Venue
Interactive Learning Environments
Publisher
Informa UK Limited
Published
2023-05-05
Volume
32
Issue
8
Pages
4613-4631
ISSN
  • 1049-4820 (ISSN-L)
  • 1744-5191
DOI
10.1080/10494820.2023.2204343
OpenAlex
W4372319207
Semantic Scholar
83209243280846b190f399468b7cf21f0ce2fe98
Cited by
143
References
13
Citations by year
2026: 35 · 2025: 51 · 2024: 48 · 2023: 9
FWCI
2.99
Type
journal article
Language
en
Open access
closed
Locations
Interactive Learning Environments (publishedVersion)
Topics
  • Physical Sciences › Computer Science › Artificial Intelligence › Speech and dialogue systems
  • Physical Sciences › Computer Science › Artificial Intelligence › Intelligent Tutoring Systems and Adaptive Learning
  • Social Sciences › Arts and Humanities › Language and Linguistics › EFL/ESL Teaching and Learning
Keywords
language learning, large language models, speaking proficiency, listening proficiency, conversation partners, learner perceptions, mobile devices

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review