The effectiveness of ChatGPT in enhancing

topical vocabulary among sophomore English

majors at Ho Chi Minh University of Banking

Huỳnh Hải Yến1, Huỳnh Lê Mỹ Trinh2, Nguyễn Phong Vũ3, Nguyễn Gia Vỹ4, Nguyễn
Ngọc Trâm5, Phạm Thái Thủy Tiên6

Keywords: ChatGPT; topical vocabulary; speaking skills; sophomore English majors.

1. Introduction
In modern days, English is one of the most popular languages spoken all over the world,
and people communicate with each other in this common language. According to Anggryadi
(2014), there is a widespread belief among speakers that studying a language may lead to
successful speaking, which is the main objective of learning English as a foreign language.
Furthermore, according to PARUPALLI, “Speaking skills are the most essential skills for all
the learners who wish to learn English to enhance their career, improve business, build
confidence levels, get better job opportunities, make public speeches, attend interviews,
participate in debates, and group….”. So without the English language speaking today, we
cannot move our day-to-day life with ease in the up-to-date world. The importance of English
speaking cannot be unexploited and discarded since the English language is the most
common language spoken all over the world.
As a second language, English is considered as difficult to learn especially in terms of
speaking skills (Hashim, Yunus, & Hashim, 2018). Recent advances in artificial intelligence
have made learning English easier. ChatGPT, developed by OpenAI and based on the GPT-3
model (2020), which laid the foundation for its conversational capabilities, is a conversational
tool that can understand and generate natural language. ChatGPT offers a simple and
interactive way to practice English. It lets users have real conversations and gives feedback
immediately. For example, studies have shown that using AI tools like ChatGPT can help

1
Huỳnh Hải Yến: Sinh viên năm 2 ngành ngôn ngữ Anh, Trường đại học Ngân Hàng thành phố Hồ
Chí Minh, MSSV: 030439230311, Gmail: huynhhaiyen2202@[Link]
2
Huỳnh Lê Mỹ Trinh: Sinh viên năm 2 ngành ngôn ngữ Anh, Trường đại học Ngân Hàng thành phố
Hồ Chí Minh, MSSV: 030439230272, Gmail: letrinhhuynh2705@[Link]
3
Nguyễn Phong Vũ: Sinh viên năm 2 ngành ngôn ngữ Anh, Trường đại học Ngân Hàng thành phố Hồ
Chí Minh, MSSV: 030439230297, Gmail: npv21102005@[Link]
4
Nguyễn Gia Vỹ: Sinh viên năm 2 ngành ngôn ngữ Anh, Trường đại học Ngân Hàng thành phố Hồ
Chí Minh, MSSV: 030439230306, Gmail: giavy050105@[Link]
5
Nguyễn Ngọc Trâm: Sinh viên năm 2 ngành ngôn ngữ Anh, Trường đại học Ngân Hàng thành phố
Hồ Chí Minh, MSSV: 030439230255, Gmail: ngoctram.3185@[Link]
6
Phạm Thái Thủy Tiên: Sinh viên năm 2 ngành ngôn ngữ Anh, Trường đại học Ngân Hàng thành phố
Hồ Chí Minh, MSSV: 030439230248, Gmail: phamthaithuytien193@[Link]
learners pick up new vocabulary faster (Smith & Johnson, 2023). Other research suggests that
this kind of tool makes it easier to improve speaking skills effectively (Chen et al., 2022).
ChatGPT has shown potential in supporting vocabulary learning by providing contextualized
and interactive input (Elgort, 2011; Lee, 2023). In speaking skills, it helps learners practice
using topic-related vocabulary in simulated dialogues, enhancing both fluency and lexical
accuracy (Wang & Vasquez, 2023).

Figure 1. HUB’s English major speaking rubrics

In English speaking assessments, vocabulary range and accuracy significantly affect
students’ performance, accounting for 25% of the total score. According to the evaluation
rubric, students who use only simple and limited vocabulary tend to score in the lower bands
(below 6), while higher scores (6–10) require a varied and accurate use of vocabulary,
including topic-specific vocabulary. Many students struggle with a lack of topic-related
vocabulary, which forces them to rely on basic, general words, and prevents them from
expressing their ideas clearly, deeply, and effectively in speaking tasks. This limits their
ability to meet the criteria for higher bands, resulting in lower speaking scores.
There have been many previous studies highlighting the effectiveness of Chat GPT in
developing and enhancing speaking skills. Research by Selvanathan et al. (2024) examined
the role of ChatGPT as a supportive tool in improving ESL learners' speaking proficiency,
demonstrating the effectiveness of ChatGPT in fostering speaking abilities and interactive
learning experiences. Besides the research around ChatGPT, AI chatbots also greatly
contribute in enhancing students’ English-speaking learning outcomes, confidence,
engagement, and motivation (Jinming Du & Daniel, 2024). Their study provides
recommendations for in-service English teachers, chatbot designers, and chatbot researchers
in the future. In addition to AI-based approaches, other research conducted by Ridayani, R.,
& Purwanto, M. B. (2024) demonstrated integrating role play and multimedia technology
significantly improved students' speaking skills. This study showed that combining role-play
and advanced technology helps students become more confident in using English, can
communicate more effectively, and show improvement in pronunciation and intonation.
There are many previous studies that have demonstrated the effectiveness of ChatGPT in
improving learners’ speaking skills. Studies focus especially on how to integrate ChatGPT to
enhance speaking skills through grammar, fluency, pronunciation, and vocabulary. The
previous research tends to emphasize common vocabulary acquisition or general speaking
proficiency rather than investigating the potential of Chat GPT for gaining thematic
vocabulary. Additionally, they also mostly highlighted the potential of Chat GPT for
speaking learning with EFL learners in general, rather than English major students.
Quality responses indicate that students perceive ChatGPT as a valuable tool for
expanding their vocabulary, citing benefits such as instant feedback and contextualized
practice. After practicing with ChatGPT, they will better understand word meanings,
collocations, and appropriate specialized contexts. Additionally, ChatGPT participants are
expected to exhibit greater speaking fluency, with fewer hesitations and increased
confidence. This study aims to evaluate how effective ChatGPT is in enhancing topical
vocabulary for speaking skills among sophomore English majors. The focus is on measuring
the extent to which ChatGPT supports learners in expanding and effectively using topic-
related vocabulary in spoken English communication. This research aims to answer the
following questions: How effective is ChatGPT in improving sophomore English majors'
topical vocabulary for speaking skills at Ho Chi Minh University of Banking?
2. Research method
Research design
A mixed-methods approach is best proposed in the context of this research. This method
combines both qualitative and quantitative research approaches to analyze sophomore
English majors at HUB who used ChatGPT to practice vocabulary and demonstrate
remarkable improvement in their speaking skill, particularly in topical vocabulary use.
The qualitative method will use interviews to gain deeper insights into their learning
experience with ChatGPT. Based on an investigation of previous research topics related to
the current one, the researcher will conduct interviews with experts in the research field to
collect detailed data. The results of these interviews will serve as a foundation for the
questionnaire. The quantitative approach focuses on collecting numerical data on students'
use of ChatGPT for vocabulary learning through a structured questionnaire. This data will
demonstrate the effectiveness of ChatGPT in enhancing students’ use of topic-specific
vocabulary. By combining these two methods, the study aims to provide both a broad
overview and a detailed understanding of how ChatGPT supports topic-specific vocabulary
learning in the context of speaking skill development.
Research participants
The study was conducted with 52 participants, including 2 lecturers from the Faculty of
Languages at Ho Chi Minh university of Banking who had expertise in the field of linguistics
and participated in an interview designed by the research team. The remaining 50 participants
were second-year students majoring in English at HUB who took part in a survey.
Procedure
This study adopts a mixed-methods design and applies two main research instruments—a
structured questionnaire through Google Forms to collect large-scale quantitative data on
students’ experiences and in-depth interviews for a deeper understanding—conducted over a
one-month period (March to April 2025) to examine the effectiveness of ChatGPT in
enhancing topical vocabulary in speaking skills among sophomore English majors at Ho Chi
Minh university of Banking.
The review of relevant literature on AI-supported vocabulary acquisition (Aljanabi &
Abdurrahman, 2023; Kim, 2023; Satar & Akcan, 2022) erves to establish a theoretical
foundation for the study. This review informs the identification of key issues and the
development of appropriate research instruments, contributing to the accurate and effective
design of the questionnaire and interview.
The interview process will involve two university lecturers to gather expert opinions on
the role of ChatGPT in enhancing vocabulary learning among sophomores. The lecturers will
be asked about their views on how AI tools like ChatGPT support vocabulary acquisition,
with a particular focus on its effectiveness in expanding a wide range of topical vocabulary
and improving specialized vocabulary for academic speaking. Additionally, the lecturers will
provide suggestions for further optimizing the integration of ChatGPT into language learning.
Interview appointments were arranged based on participants' preferences, and each session
lasted about 10–15 minutes. All interviews were conducted either via video conferencing or
face-to-face. Participants provided informed consent before the interviews, which were
audio-recorded, transcribed exactly. These interviews aim to complement the data collected
from the student questionnaires, offering valuable insights into the impact of AI on
vocabulary development.
The survey was distributed to 100 participants within 1 week. The data from the interview
and the experts’ sharing helped create the survey questionnaire. The questionnaire structure
contains 18 items are closed-ended questions structured into three sections: the first gathers
general information such as gender, proficency level and the frequency of ChatGPT usage;
the second uses 5-point Likert scale items (1 = strongly disagree to 5 = strongly agree)
measuring perceived effectiveness of ChatGPT in learning and the last part evaluates their
attitudes toward the topic. The data collected from the questionnaire will be compiled and
analyzed using Microsoft Excel. The findings will assist in identifying the effectiveness of
ChatGPT in learning topical vocabulary of students.
Data collection and data analysis
 This part aims to analyze data collection for a mixed-method research study. An online
questionnaire was created and shared with sophomores via emails for 1 week. Their answers
were kept anonymous because they didn’t need to provide their names. The survey was
conducted over one month, from March to April. In addition to the questionnaire, we
conducted interviews with two English lecturers from Ho Chi Minh University of Banking.
These interviews aimed to collect their perceptions of the effectiveness of ChatGPT in
enhancing topical vocabulary among sophomore English majors. This qualitative approach
provided a deeper understanding of the educators' views and experiences obtained from the
survey.
The data analysis in this study was conducted using a mixed-methods approach,
integrating both quantitative and qualitative data to ensure the comprehensive
of effectiveness of ChatGPT in enhancing topical vocabulary in speaking skills among
sophomore English majors at HUB. Quantitative data were obtained from 100 students to
analyze the engagement of topical vocabulary learning in ChatGPT. The results were
converted to Excel, a statistical tool, and calculated the descriptive statistics, researchers then
ran the inferential statistics to test the research’s hypotheses. Lastly, present the final results
of quantitative data in tables. In parallel, qualitative data were collected through other
research papers, previous studies and data sources from specialists along with semi-structured
interviews. To get a deeper grip onto the data, coding is widely applied, and researchers find
patterns, keywords, and recurring themes, then group similar codes into categories or themes
manually. By integrating the two sets of data, researchers observed the quantitative and
qualitative findings and used this integration to provide a more complete understanding of
both the research problem and the questionnaire. Finally, clearly explain why and how the
data were integrated in the research write-up.

Reference
Lee, J. (2023). Contextualized vocabulary instruction through AI-based tools. Journal of Educational
Technology, 20(3), 200–215.
PARUPALLI, S. R. (2019). Significance of English Communication Skills in the Modern Era.
Journal for Research Scholars and Professionals of English Language Teaching, 3(26), 1–7.
Purohit, B., Saraswati, K., & Devi, A. (2023). Mixed methods research design in educational studies:
Approaches and best practices. International Journal of Educational Research Methods, 5(1),
12–25.
Ridayani, R., & Purwanto, M. B. (2024). The integration of multimedia and role-play to develop
students’ speaking skills. International Journal of Language Learning, 9(1), 34–48.
Selvanathan, M., Tan, H. L., & Kumar, D. (2024). Enhancing ESL learners’ speaking proficiency
using ChatGPT: A study on interactive learning. Asian Journal of Education and e-Learning,
12(2), 89–101.
Smith, A., & Johnson, L. (2023). Accelerated vocabulary acquisition using AI-powered tools.
Language Technology & Learning, 11(4), 78–93.
Torky, S. A. (2006). The effectiveness of a task-based instruction program in developing the English
language speaking skills of secondary stage students. [Unpublished doctoral dissertation]. Ain
Shams University.
Wang, X., & Vasquez, C. (2023). AI-supported dialogue systems for second language speaking
practice: Affordances and challenges. Computer Assisted Language Learning, 36(5), 413–432.
Zhang, Y. (2009). Reading to Speak: Integrating Oral Communication Skills. English Teaching
Forum, 47(1), 32–34.

Questionnaire
Survey on the effectiveness of ChatGPT in enhancing topical
vocabulary in speaking skills among sophomore English majors in
HUB
Hi everyone! We are a group of second-year students of the Faculty of Foreign Languages
majoring in English Language at Ho Chi Minh University of Banking.
Our group comprises 6 members: Huynh Hai Yen, Huynh Le My Trinh, Nguyen Phong Vu,
Nguyen Ngoc Tram, Pham Thai Thuy Tien, and Nguyen Phong Vu under the guidance of
Ms. Nguyen Thi Bich Phuong, a lecturer at the Faculty of Foreign Languages, HUB. We are
conducting a research project on "The effectiveness of ChatGPT in enhancing topical
vocabulary in speaking skills among sophomore English majors in HUB". This study aims to
demonstrate improvement in their speaking skills, especially the flexibility of using topical
vocabulary. Therefore, this survey is designed to collect data/information for the research
article.
Your cooperation by answering these questions will contribute significantly to the success of
the research in general and our team in particular. We hope you will answer these questions
carefully and honestly with your personal views. We commit that all data collected will be
used strictly for research purposes and will not be used for any other purpose.
PART 1: PERSONAL INFORMATION
In this section, we collect your basic information for the analysis and research of the thesis. The
information you provide will be kept confidential and used solely for research purposes. Please
provide accurate information for the following fields:
1. What is your gender identity? □ Male□ Female
2. Which year are you in?
□ Freshmen □ Sophomore □ Junior □ Senior
3. How long have you been using ChatGPT?
□ Less than 1 month
□ 1-3 months
□ 3-6 months
□ More than 6 months
4. How frequently do you use ChatGPT for learning topical vocabulary per day?
□ Less than 30 minutes
□ 1-2 hours
□ 2-3 hours
□ More than 3 hours
PART 2: THE EFFECTIVENESS OF CHATGPT IN ENHANCING TOPICAL
VOCABULARY IN SPEAKING SKILLS
Please rate the following statements on a 5-point Likert scale (1 = Strongly Disagree, 2 =
Disagree, 3 = Neutral, 4 = Agree, 5 = Strongly Agree).

1. ChatGPT helps me learn vocabulary across a wide 1 2 3 4 5

range of topics. □ □ □ □ □
2. ChatGPT helps me expand the amount of vocabulary 1 2 3 4 5
on a specific topic. □ □ □ □ □
3. ChatGPT gives me synonyms of topical vocabulary 1 2 3 4 5
at different stages of levels. □ □ □ □ □
4. I can learn a wide range of related words and 1 2 3 4 5
expressions within one topic using ChatGPT. □ □ □ □ □
5. The amount of vocabulary I learn from ChatGPT for 1 2 3 4 5
each topic is greater than that from other methods. □ □ □ □ □
6. The topic-based vocabulary suggested by ChatGPT 1 2 3 4 5
is useful for academic speaking. □ □ □ □ □
7. ChatGPT can adjust the vocabulary difficulty 1 2 3 4 5
depending on my level or request. □ □ □ □ □
8. ChatGPT provides me with contextualized examples 1 2 3 4 5
of new vocabulary that are understandable. □ □ □ □ □
PART 3: STUDENTS' ATTITUDES TOWARD THE RESEARCH TOPIC
Please rate the following statements on a 5-point Likert scale (1 = Strongly Disagree, 2 =
Disagree, 3 = Neutral, 4 = Agree, 5 = Strongly Agree).

1. I find the topic “How ChatGPT enhances topical 1 2 3 4 5

vocabulary in speaking” interesting. □ □ □ □ □
2. This topic is practically useful for English learners 1 2 3 4 5
like me. □ □ □ □ □
3. I believe ChatGPT can support vocabulary 1 2 3 4 5
development for speaking skills. □ □ □ □ □
4. This research topic fits well with the needs of 1 2 3 4 5
students today. □ □ □ □ □
5. I am willing to apply ChatGPT in learning 1 2 3 4 5
vocabulary after knowing this research. □ □ □ □ □
6. I think the results of this study can benefit both 1 2 3 4 5
students and teachers. □ □ □ □ □
Interview
Including 2 open-ended questions:
1. Do you encourage students to expand their vocabulary on topics by using ChatGPT?
2. Do you find that students use ChatGPT effectively to develop topical vocabulary
during their learning process?