Made byBobr AI

ChatGPT in EFL Writing: Impact on Brainstorming & Anxiety

Research on the effects of ChatGPT-assisted brainstorming on idea generation and writing anxiety among EFL Master's students, featuring the SCAFFOLD model.

#chatgpt#efl-writing#brainstorming-ai#writing-anxiety#educational-technology#academic-research#scaffold-model#ai-in-education
Watch
Pitch
Research Questions & Hypotheses
Section 04 — Research Questions & Hypotheses

Research Questions

RQ1.
To what extent does ChatGPT-assisted brainstorming enhance EFL Master's students' idea generation?
RQ2.
How does ChatGPT-assisted brainstorming influence EFL students' writing anxiety?
RQ3.
What are the perceptions of EFL Master's students regarding ChatGPT as a brainstorming assistant?

Research Hypotheses

RH1(0):
ChatGPT-assisted brainstorming has no significant effect on EFL Master's students' idea generation.
RH1(1):
ChatGPT-assisted brainstorming significantly enhances EFL Master's students' idea generation.
RH2(0):
ChatGPT-assisted brainstorming has no significant effect on EFL Master's students' writing anxiety.
RH2(1):
ChatGPT-assisted brainstorming significantly reduces EFL Master's students' writing anxiety.
RH3:
EFL Master's students display positive perceptions toward ChatGPT as an AI-assisted brainstorming tool.
Made byBobr AI
Methodology
Section 05

Methodology

01
01

PARADIGM

Pragmatic paradigm — concerned with practical applications and solutions to real-world problems; justifies mixed-methods approach.
02
02

APPROACH

Explanatory Sequential Mixed-Methods — quantitative data collected and analyzed first, then qualitative data explains or interprets the results.
03
03

SAMPLE

N = 28 first-year Master's EFL students at Biskra University (convenience sampling).

+ 6 students purposively selected exclusively for focus group interviews.
04
04

INSTRUMENTS

  • Adapted SLWAI: 22-item Likert scale (somatic, cognitive & avoidance anxiety).
  • Idea Generation Tasks: Task A & B — timed brainstorming with analytic rubric.
  • Focus Group Interview: Semi-structured format, 6 participants.
05
05

DATA ANALYSIS

Quantitative (IBM SPSS 27):
Descriptive stats, paired-samples t-test, Cohen's d effect size, normality tests.
Qualitative (MAXQDA):
Thematic analysis (Braun & Clarke, 2022).
06
06

INTERVENTION

5-Session ChatGPT Training:
SCAFFOLD model (8 ethical steps) + structured prompting strategies.
Pre-test → Intervention → Post-test → FGD
Made byBobr AI
Results: Idea Generation
Section 06
Component Pre-Test Mean (SD) Post-Test Mean (SD)
Quantity 1.71 (0.76) 2.00 (0.72)
Relevance 2.54 (0.58) 2.18 (0.72)
Diversity 2.00 (0.77) 2.32 (0.77)
Total Task Score 6.25 (1.80) 6.50 (1.95)
Components range 1–3; Total range 3–9; N = 28
•
Paired-samples t-test:
t(27) = 0.72, p = .478 (not significant)
•
Effect Size:
Cohen's d = 0.14 (very small)
•
Performance shift:
High performance ↑ from 28.6% to 42.9%
" ChatGPT-assisted brainstorming had a positive but non-significant statistical effect on idea generation. Qualitative findings, however, confirmed perceived improvements in quantity, diversity, and idea organization — attributed to ethical scaffolded use. "
Null hypothesis RH1(0) retained.
Made byBobr AI
Results: Writing Anxiety
Section 07

Results: Writing Anxiety (SLWAI)

Subscale Pre-Test M (per item) Post-Test M (per item) Change
Somatic Anxiety (10 items) 2.77 2.37 ↓ −0.40
Cognitive Anxiety (7 items) 3.10 2.60 ↓ −0.50 (largest reduction)
Avoidance Behavior (5 items) 2.49 2.40 ↓ −0.09
Total SLWAI 61.86 (SD=14.10) 53.89 (SD=14.07) ↓ −7.97

Anxiety Level Distribution

Low anxiety: 21.4% → 46.4%
(more than doubled)
Moderate anxiety: 42.9% → 28.6%
High anxiety: 35.7% → 25.0%
Pre-test dominant factor: Cognitive anxiety was the highest source of stress, impacting 70% of the sample prior to intervention.

Significant Findings

★
Paired-samples t-test: t(27) = −3.491, p = .002
(statistically significant)
★
Effect Size: Cohen's d = −0.66 (medium effect)
★
Null hypothesis RH2(0) REJECTED — Alternative RH2(1) supported
★
Cronbach's alpha reliability:
pre-test α = .894 (Good)  ·  post-test α = .914 (Excellent)
"ChatGPT-assisted brainstorming significantly reduced EFL Master's students' writing anxiety, with the largest reduction in cognitive anxiety — the most prevalent dimension."
Made byBobr AI
Results: Focus Group
SECTION 08

Results: Focus Group Interview

Sample: 6 purposively selected participants • 45-min audio recorded session • Analyzed via MAXQDA
THEME 1

ChatGPT as a Facilitator of Idea Generation

  • Participants praised ChatGPT for increasing both quantity and quality of ideas
  • Helped them "think out of the box" and explore new perspectives
  • Supported idea organization and logical structure
"So as if ChatGPT show you the right path of collecting data, even in quantity or quality."
Theme 2

Reduction of Writing Anxiety & Increased Confidence

  • ChatGPT significantly reduced stress, boosted motivation and confidence
  • Especially effective with SCAFFOLD model and prompting strategies
"Before using ChatGPT I felt stressed, unmotivated... However, with using ChatGPT I feel more confident, motivated... It reduced my anxiety about writing."
Theme 3

Ethical Use through Structured Scaffolding & Guided Prompting

  • SCAFFOLD model and prompting strategies were key to ethical use
  • Participants distinguished between brainstorming assistant vs. ghostwriter
"After the experience and after using brainstorming and scaffolding model and the prompts, I changed my mind. Because using the prompts, it will give you questions that guide you, not answer for you."
Theme 4

Challenges and Limitations of ChatGPT Use

  • Risk of over-reliance and potential creativity loss
  • Concerns about plagiarism under time pressure
  • Accessibility issues (internet dependency)
"I was afraid from being lazy, just using ChatGPT, not depending on my own thinking, my critical thinking... Do not give your creativity, your brilliant ideas."
Made byBobr AI
Recommendations
Section 09 — Recommendations
🏛
POLICYMAKERS
Clearly define the permissible use, misuse, and abuse of AI tools in educational contexts. Encourage and ensure AI accessibility within universities, and develop institutional frameworks that support ethical integration — especially in language learning.
📚
CURRICULUM DESIGNERS
Adapt educational programs to modern needs by developing and integrating AI-assisted writing initiatives. Ensure smooth, structured incorporation of ChatGPT and similar tools into the curriculum to facilitate learners' language learning process.
👩‍🏫
TEACHERS
Shift perspectives on AI — view ChatGPT less as a competitor and more as a collaborator. Adopt welcoming stances toward AI tools that foster both cognitive and emotional engagement in writing classrooms. Provide training in ethical AI use.
🎓
STUDENTS
Avoid over-reliance on AI. Use ChatGPT as a thinking partner and scaffolding tool — not a shortcut. Apply the SCAFFOLD model and structured prompts to maintain critical thinking, creativity, and academic integrity.
Made byBobr AI
Conclusion
Section 10

Summary of Findings

•
RQ1 — Idea Generation: ChatGPT-assisted brainstorming had a positive but non-significant statistical effect (t(27) = 0.72, p = .478; d = 0.14). Qualitative findings, however, revealed perceived improvements in quantity, diversity, and idea organization via the SCAFFOLD model.
•
RQ2 — Writing Anxiety: A statistically significant reduction was found (t(27) = −3.491, p = .002; d = −0.66, medium effect). All SLWAI subscales decreased, with cognitive anxiety showing the largest drop.
•
RQ3 — Student Perceptions: Overwhelmingly positive perceptions of ChatGPT as a brainstorming assistant. SCAFFOLD model and prompting strategies were central to ethical and effective use.
This dissertation is divided into two main chapters: Chapter One provides the theoretical framework covering EFL writing, brainstorming, writing anxiety, and AI in language learning; Chapter Two presents the methodological choices, fieldwork, data analysis, discussion, and synthesis of findings.
"When implemented properly, AI tools — specifically ChatGPT — can meaningfully support EFL learners by enhancing idea generation and reducing writing anxiety, while maintaining ethical and academic integrity."
Made byBobr AI
Mohamed Kheider University – Biskra
Faculty of Letters and Foreign Languages
Dept. of Language & English Literature
Academic Year: 2025/2026
Master's Dissertation Defense

Thank You

Questions & Comments are Welcome
Submitted by:
Ali MJENNAH
Supervised by:
Dr. Messaouda BENDAHMANE
Keywords:
AI
Brainstorming
ChatGPT
Foreign Language Writing Anxiety
Idea Generation
SCAFFOLD Model
EFL
Mixed Methods
Mohamed Kheider University of Biskra — 2025/2026
Made byBobr AI
Bobr AI

DESIGNER-MADE
PRESENTATION,
GENERATED FROM
YOUR PROMPT

Create your own professional slide deck with real images, data charts, and unique design in under a minute.

Generate For Free

ChatGPT in EFL Writing: Impact on Brainstorming & Anxiety

Research on the effects of ChatGPT-assisted brainstorming on idea generation and writing anxiety among EFL Master's students, featuring the SCAFFOLD model.

Section 04 — Research Questions & Hypotheses

Research Questions & Hypotheses

Research Questions

Research Hypotheses

To what extent does ChatGPT-assisted brainstorming enhance EFL Master's students' idea generation?

How does ChatGPT-assisted brainstorming influence EFL students' writing anxiety?

What are the perceptions of EFL Master's students regarding ChatGPT as a brainstorming assistant?

ChatGPT-assisted brainstorming has no significant effect on EFL Master's students' idea generation.

ChatGPT-assisted brainstorming significantly enhances EFL Master's students' idea generation.

ChatGPT-assisted brainstorming has no significant effect on EFL Master's students' writing anxiety.

ChatGPT-assisted brainstorming significantly reduces EFL Master's students' writing anxiety.

EFL Master's students display positive perceptions toward ChatGPT as an AI-assisted brainstorming tool.

Section 05

Methodology

Methodology

PARADIGM

<div><strong style="color: #1a4a2e; font-weight: 600;">Pragmatic paradigm</strong> &mdash; concerned with practical applications and solutions to real-world problems; justifies mixed-methods approach.</div>

APPROACH

<div><strong style="color: #1a4a2e; font-weight: 600;">Explanatory Sequential Mixed-Methods</strong> &mdash; quantitative data collected and analyzed first, then qualitative data explains or interprets the results.</div>

SAMPLE

<div><strong style="color: #1a4a2e; font-weight: 600;">N = 28</strong> first-year Master's EFL students at Biskra University (convenience sampling).</div><br><div><strong style="color: #1a4a2e; font-weight: 600;">+ 6 students</strong> purposively selected exclusively for focus group interviews.</div>

INSTRUMENTS

<ul style="margin: 0; padding-left: 20px; display: flex; flex-direction: column; gap: 8px;"><li><strong style="color: #1a4a2e; font-weight: 600;">Adapted SLWAI:</strong> 22-item Likert scale (somatic, cognitive & avoidance anxiety).</li><li><strong style="color: #1a4a2e; font-weight: 600;">Idea Generation Tasks:</strong> Task A & B &mdash; timed brainstorming with analytic rubric.</li><li><strong style="color: #1a4a2e; font-weight: 600;">Focus Group Interview:</strong> Semi-structured format, 6 participants.</li></ul>

DATA ANALYSIS

<div style="margin-bottom: 15px;"><strong style="color: #1a4a2e; font-weight: 600;">Quantitative (IBM SPSS 27):</strong><br>Descriptive stats, paired-samples t-test, Cohen's d effect size, normality tests.</div><div><strong style="color: #1a4a2e; font-weight: 600;">Qualitative (MAXQDA):</strong><br>Thematic analysis (Braun & Clarke, 2022).</div>

INTERVENTION

<div><strong style="color: #1a4a2e; font-weight: 600;">5-Session ChatGPT Training:</strong><br>SCAFFOLD model (8 ethical steps) + structured prompting strategies.</div><div style="background: rgba(26, 74, 46, 0.06); padding: 12px 15px; border-radius: 6px; display: flex; align-items: center; justify-content: space-between; width: 100%; box-sizing: border-box; font-size: 16px; font-weight: 600; color: #1a4a2e; margin-top: auto; border: 1px solid rgba(26, 74, 46, 0.15);"><span>Pre-test</span> &rarr; <span>Intervention</span> &rarr; <span>Post-test</span> &rarr; <span>FGD</span></div>

Results: Idea Generation

Section 06

Component

Pre-Test Mean (SD)

Post-Test Mean (SD)

Quantity

1.71 (0.76)

2.00 (0.72)

Relevance

2.54 (0.58)

2.18 (0.72)

Diversity

2.00 (0.77)

2.32 (0.77)

Total Task Score

6.25 (1.80)

6.50 (1.95)

Components range 1–3; Total range 3–9; N = 28

t(27) = 0.72, p = .478 (not significant)

Cohen's d = 0.14 (very small)

High performance ↑ from 28.6% to 42.9%

ChatGPT-assisted brainstorming had a positive but non-significant statistical effect on idea generation. Qualitative findings, however, confirmed perceived improvements in quantity, diversity, and idea organization — attributed to ethical scaffolded use.

Null hypothesis RH1(0) retained.

Results: Writing Anxiety

Section 07

Results: Writing Anxiety (SLWAI)

ChatGPT-assisted brainstorming significantly reduced EFL Master's students' writing anxiety, with the largest reduction in cognitive anxiety — the most prevalent dimension.

Results: Focus Group

SECTION 08

Results: Focus Group Interview

6 purposively selected participants

45-min audio recorded session

Analyzed via MAXQDA

THEME 1

ChatGPT as a Facilitator of Idea Generation

Participants praised ChatGPT for increasing both quantity and quality of ideas

Helped them "think out of the box" and explore new perspectives

Supported idea organization and logical structure

So as if ChatGPT show you the right path of collecting data, even in quantity or quality.

Theme 2

Reduction of Writing Anxiety & Increased Confidence

ChatGPT significantly reduced stress, boosted motivation and confidence

Especially effective with SCAFFOLD model and prompting strategies

Before using ChatGPT I felt stressed, unmotivated... However, with using ChatGPT I feel more confident, motivated... It reduced my anxiety about writing.

Theme 3

Ethical Use through Structured Scaffolding & Guided Prompting

SCAFFOLD model and prompting strategies were key to ethical use

Participants distinguished between brainstorming assistant vs. ghostwriter

After the experience and after using brainstorming and scaffolding model and the prompts, I changed my mind. Because using the prompts, it will give you questions that guide you, not answer for you.

Theme 4

Challenges and Limitations of ChatGPT Use

Risk of over-reliance and potential creativity loss

Concerns about plagiarism under time pressure

Accessibility issues (internet dependency)

I was afraid from being lazy, just using ChatGPT, not depending on my own thinking, my critical thinking... Do not give your creativity, your brilliant ideas.

Section 09 — Recommendations

Recommendations

🏛

POLICYMAKERS

Clearly define the permissible use, misuse, and abuse of AI tools in educational contexts. Encourage and ensure AI accessibility within universities, and develop institutional frameworks that support ethical integration — especially in language learning.

📚

CURRICULUM DESIGNERS

Adapt educational programs to modern needs by developing and integrating AI-assisted writing initiatives. Ensure smooth, structured incorporation of ChatGPT and similar tools into the curriculum to facilitate learners' language learning process.

👩‍🏫

TEACHERS

Shift perspectives on AI — view ChatGPT less as a competitor and more as a collaborator. Adopt welcoming stances toward AI tools that foster both cognitive and emotional engagement in writing classrooms. Provide training in ethical AI use.

🎓

STUDENTS

Avoid over-reliance on AI. Use ChatGPT as a thinking partner and scaffolding tool — not a shortcut. Apply the SCAFFOLD model and structured prompts to maintain critical thinking, creativity, and academic integrity.

Conclusion

Section 10

Summary of Findings

RQ1 — Idea Generation:

ChatGPT-assisted brainstorming had a positive but non-significant statistical effect (t(27) = 0.72, p = .478; d = 0.14). Qualitative findings, however, revealed perceived improvements in quantity, diversity, and idea organization via the SCAFFOLD model.

RQ2 — Writing Anxiety:

A statistically significant reduction was found (t(27) = −3.491, p = .002; d = −0.66, medium effect). All SLWAI subscales decreased, with cognitive anxiety showing the largest drop.

RQ3 — Student Perceptions:

Overwhelmingly positive perceptions of ChatGPT as a brainstorming assistant. SCAFFOLD model and prompting strategies were central to ethical and effective use.

This dissertation is divided into two main chapters: Chapter One provides the theoretical framework covering EFL writing, brainstorming, writing anxiety, and AI in language learning; Chapter Two presents the methodological choices, fieldwork, data analysis, discussion, and synthesis of findings.

When implemented properly, AI tools — specifically ChatGPT — can meaningfully support EFL learners by enhancing idea generation and reducing writing anxiety, while maintaining ethical and academic integrity.

Mohamed Kheider University – Biskra

Faculty of Letters and Foreign Languages

Dept. of Language & English Literature

Academic Year: 2025/2026

Master's Dissertation Defense

Thank You

Questions & Comments are Welcome

Submitted by:

Ali MJENNAH

Supervised by:

Dr. Messaouda BENDAHMANE

Keywords:

AI

Brainstorming

ChatGPT

Foreign Language Writing Anxiety

Idea Generation

SCAFFOLD Model

EFL

Mixed Methods

Mohamed Kheider University of Biskra — 2025/2026