AI evidence extraction

Can ChatGPT Scaffold Postgraduate Students’ Thesis Proposals? Voices of EFL Teachers and Students

Yudhi Arifani, Khoirul Anwar, Nirwanto Maruf, Rina Wahyu Setyaningrum, Masduki Masduki, Sahiruddin Sahiruddin · International Journal of Learning Teaching and Educational Research · 2026

AI-generated evidence extraction, verified across multiple analytical personas. Not a substitute for the peer-reviewed original.

8/10

Relevance

1/4

Quality (LMQS)

Evidence

Citations

0.00

FWCI

This is an AI analysis. Read the peer-reviewed original at the publisher: https://doi.org/10.26803/ijlter.25.2.40

Methodology & findings

Study design

Mixed-methods design combining quantitative questionnaire survey (120 postgraduate students) with Mann-Whitney U test analysis, and qualitative semi-structured interviews (24 students, 12 instructors) analyzed through five-phase thematic analysis.

Sample

N = 156, 6 groups

Primary method

Mann-Whitney U test for analyzing quantitative data and determining differences between students' and teachers' perspectives; five-phase thematic analysis for qualitative results; questionnaire based on moves analysis framework and technology acceptance model.

Main result

The study found significant differences in perspectives between students and teachers regarding ChatGPT's utility. Specifically, "Although students had generally positive opinions of ChatGPT, teachers were concerned about its limitations in helping students identify research gaps, synthesize previous studies, and effectively plan methodologies." Additionally, "most students relied too heavily on generic artificial intelligence outputs and they applied little critical thinking or contextual adaptation."

Reports effect sizes.

Research paradigm

Mixed-methods (pragmatist/interpretivist)

Author conclusions

The authors conclude that a mixed-methods, multi-perspective approach is necessary to understand ChatGPT integration in academic writing. The study demonstrates "significant differences in perspectives" between students and teachers, with the qualitative findings revealing that "Patterns in topic development, research gap formulation, literature synthesis, theoretical framing, and methodological planning were identified through a five-phase thematic analysis." This suggests limitations in students' critical application of AI tools.

Risk of bias

Potential selection bias in interview participants (24 of 120 survey respondents); Single-perspective approach noted as limitation in related studies; No mention of blinding or randomization; Unequal comparison groups (students vs. instructors with different sample sizes); Selection bias: Voluntary participation in interviews may skew responses toward more engaged or motivated students; Social desirability bias: Students and teachers may provide responses they perceive as appropriate rather than authentic; Interviewer bias: Open-ended interview questions may be subject to interviewer interpretation; Confounding variables: No mention of controlling for prior experience with AI or writing ability; Potential selection bias in interview participants (24 of 120 students self-selected); Single-institution or program setting not specified; Instructor bias not measured or controlled; Self-reported perceptions via questionnaire and interviews subject to social desirability bias

Open questions raised

The authors identify the need for studies that adopt multi-perspective approaches and have sufficiently articulated theoretical and methodological grounding when examining ChatGPT's role in academic writing and thesis proposal development.
The authors note that "Studies on ChatGPT and academic writing often adopt a single-perspective approach, and lack sufficiently articulated theoretical and methodological grounding," suggesting need for more integrated, multi-stakeholder approaches to understanding AI in academic writing.
The abstract indicates the study addresses the gap that "Studies on ChatGPT and academic writing often adopt a single-perspective approach, and lack sufficiently articulated theoretical and methodological grounding." Future research directions include examining how to improve critical thinking in students' use of AI scaffolding tools.

Data: not_statedCode: not_statedExtracted from: pdfAgreement 62%

Explore related topics

AI Manuscript Writing Academic Writing Mixed Methods