Useful but Not Always Right: Engineering Students’ Use of ChatGPT for English Language Learning in a Non-Major Classroom in Bangladesh

Generative AI tools such as ChatGPT are now part of how many university students study, but we still know little about how students outside English departments use them to learn English. This exploratory survey looked at engineering students taking a non-major English course at a technical university in northern Bangladesh. Sixty-three students answered an online questionnaire with closed and open-ended items on how they use ChatGPT, what benefits they see and what problems they face. Use was very common: 96.8% had used ChatGPT, 55.6% used it daily and 54.0% used it often or always for English-related tasks, mainly for writing help, assignment preparation and generating ideas. Most students agreed that ChatGPT is easy to use (76.2%), gives useful explanations (73.0%) and saves time (71.4%). At the same time, 76.2% said its answers are not always accurate or relevant, 74.6% wanted more guidance on using it, and 66.7% reported that poor internet access limits regular use. More than half (57.1%) held both views together, finding the explanations useful and the answers unreliable. Students who used ChatGPT more often for English reported greater benefits (ρ = .41) and more over-reliance (ρ = .38), but they were no more likely to notice unreliable output (ρ = −.14). Only 39.7% preferred asking ChatGPT to asking a teacher. Open-ended answers pointed to wrong answers, language that is too advanced, and difficulty writing prompts. Because the sample is small, drawn from one institution and based on self-report, the findings describe students’ experience and not measured gains in proficiency. They suggest that these students need help in judging ChatGPT’s output more than they need encouragement to use it.

Authors

Institutions

Publication Details

Journal
International Journal of Arts and Humanities
Published
2026-11-19
DOI
https://doi.org/10.61424/ijah.v4i4.1149
Primary Topic
Artificial Intelligence in Education
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Useful but Not Always Right: Engineering Students’ Use of ChatGPT for English Language Learning in a Non-Major Classroom in Bangladesh

Nishat Tasneem, Jannatul Ferdous, Riyad Ahamed
International Journal of Arts and Humanities
Artificial Intelligence in Education
article

Useful but Not Always Right: Engineering Students’ Use of ChatGPT for English Language Learning in a Non-Major Classroom in Bangladesh

Nishat Tasneem, Jannatul Ferdous, Riyad Ahamed
article en

Abstract

Generative AI tools such as ChatGPT are now part of how many university students study, but we still know little about how students outside English departments use them to learn English. This exploratory survey looked at engineering students taking a non-major English course at a technical university in northern Bangladesh. Sixty-three students answered an online questionnaire with closed and open-ended items on how they use ChatGPT, what benefits they see and what problems they face. Use was very common: 96.8% had used ChatGPT, 55.6% used it daily and 54.0% used it often or always for English-related tasks, mainly for writing help, assignment preparation and generating ideas. Most students agreed that ChatGPT is easy to use (76.2%), gives useful explanations (73.0%) and saves time (71.4%). At the same time, 76.2% said its answers are not always accurate or relevant, 74.6% wanted more guidance on using it, and 66.7% reported that poor internet access limits regular use. More than half (57.1%) held both views together, finding the explanations useful and the answers unreliable. Students who used ChatGPT more often for English reported greater benefits (ρ = .41) and more over-reliance (ρ = .38), but they were no more likely to notice unreliable output (ρ = −.14). Only 39.7% preferred asking ChatGPT to asking a teacher. Open-ended answers pointed to wrong answers, language that is too advanced, and difficulty writing prompts. Because the sample is small, drawn from one institution and based on self-report, the findings describe students’ experience and not measured gains in proficiency. They suggest that these students need help in judging ChatGPT’s output more than they need encouragement to use it.

International Journal of Arts and HumanitiesVol. 4(4)
Bangladesh Army University of Science and Technology (BD)
Artificial Intelligence in Education
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon