New Research Reveals Impact of Natural Language Processing on EFL Speaking Skills

Strategic Insights -

Researchers at Shanghai International Studies University have conducted a study to examine the effect of Natural Language Processing (NLP) tools on EFL speaking skills. The study, which involved 436 EFL learners, found that a chatbot treatment produced the highest gains in adaptability, accuracy, and fluency. The research suggests that chatbots uniquely optimize sociocultural learning mechanisms, enabling sustained fluency gains. This breakthrough has significant implications for language instruction and the use of AI in education.

Key Takeaways:

  • The study found that a chatbot treatment produced the highest gains in adaptability, accuracy, and fluency, with significant improvements in all areas.
  • The chatbot treatment involved 20 role-play sessions via ChatGPT, and the results showed that learners who used the chatbot had a mean score of 85.50 in adaptability, 84.24 in accuracy, and 85.04 in fluency.
  • The study also found that the Machine Translation Treatment and Automatic Summarization Treatment had less significant effects on EFL speaking skills, but still produced notable improvements.
  • The researchers concluded that chatbots uniquely optimize sociocultural learning mechanisms, enabling sustained fluency gains.
  • Jing Zhang, Lipei Li, Jingyi Luo, and Qiaoyun Liao are among the researchers who contributed to this study.
  • The study has significant implications for language instruction and the use of AI in education, highlighting the potential of NLP tools to enhance EFL speaking skills.

Statistics:

  • The study involved 436 EFL learners, who were randomly assigned to one of four groups: chatbot treatment, machine translation treatment, automatic summarization treatment, or control group.
  • The chatbot treatment involved 20 role-play sessions via ChatGPT, and the results showed that learners who used the chatbot had a mean score of 85.50 in adaptability, 84.24 in accuracy, and 85.04 in fluency.
  • The study found significant improvements in all areas, with a delta of +40.25 in adaptability, +43.93 in accuracy, and +42.54 in fluency.

Sources:

  • The Influence of Natural Language Processing On Efl Speaking Skills: Investigating Learner Adaptability, Language Accuracy, and Fluency. Journal of Educational Computing Research, 2025. (Peer-reviewed)
  • Journal of Educational Computing Research can be contacted at: Sage Publications Inc, 2455 Teller Rd, Thousand Oaks, CA 91320, USA.
  • Shanghai International Studies University, Inst Language Sci, 1550 Wenxiang Rd, Shanghai, People's Republic of China.