Jaroslaw Kriukow

ChatGPT for Thesis Writing

Practical Guide and Prompts for Every Stage of Dissertation Writing
How to cite this source?

In APA style, follow this format:

Kriukow, J. (2025). The Scholar’s Guide to AI-Assisted Thematic

Analysis (2nd ed.) [PDF]. https://payhip.com/b/jGE46
Contents

Introduction ............................................................................................................... 1
How This Book is Organised ....................................................................................... 2
Customizing ChatGPT before we start ....................................................................... 2
1. Planning and Idea Development ............................................................................ 6
2. Research Design ..................................................................................................... 7
Choosing methodology and identifying philosophical assumptions ..................... 7
Justifying your research approach ......................................................................... 8
Sample size ............................................................................................................. 9
Developing data collection tools ............................................................................ 9
Ethics and consent ............................................................................................... 11
Pilot testing .......................................................................................................... 11
3. Literature Work and Source Mapping .................................................................. 12
4. Coding and Data Analysis ..................................................................................... 13
Initial coding ......................................................................................................... 13
Focused (axial) coding and grouping.................................................................... 14
Developing themes .............................................................................................. 15
5. Writing and Drafting ............................................................................................. 16
Abstracts and summaries ..................................................................................... 16
Literature Review ................................................................................................. 16
Methodology........................................................................................................ 17
Results .................................................................................................................. 17
Discussion............................................................................................................. 18
Implications and Contributions ............................................................................ 18
Visualising the results .......................................................................................... 18
6. Polishing, Referencing, and Formatting ............................................................... 19
Proofreading ........................................................................................................ 19
Referencing and citation style .............................................................................. 20
Other formatting tasks ......................................................................................... 20
7. Critical Reflection and Viva Prep .......................................................................... 21
Conclusion ................................................................................................................ 22
Introduction

The world of academic work has changed dramatically, and at the

heart of this change is the rise of AI tools.

Not long ago, I genuinely believed there was no way AI could do

anything useful and meaningful in academic research - let alone
analyse data or help with
writing.

Since then, I’ve gone

from occasionally playing
with AI and feeling
vaguely guilty about it, to
fully recognising its
potential. Not just as a
convenient assistant, but
as a powerful tool that
makes me more
productive and efficient.
And I know I’m not the only one who’s made that shift.

AI is everywhere now. And yes, that includes academic assignments.

That’s what this book is about: using ChatGPT (or similar tools) to
support the process of writing a dissertation or thesis. I’m not here to
tell you what’s right or wrong. If you’ve followed me for a while, you’ll
know that the last thing I’d ever suggest is outsourcing your entire
project to AI. I enjoy the process of learning, the creativity that comes
with analysis, and the satisfaction of building something meaningful
from scratch.

1
But I’m also not naïve. I know I’m part of a generation that is slowly,
inevitably becoming the past. The new generation is faster, more
creative, more tech-savvy, and - whether we like it or not - already
way ahead in adapting to the AI revolution. And yet, too many of us
are still stuck trying to argue that it’s somehow wrong to use AI,
instead of asking the better question: Where do we go from here?

In the following pages, I’ll share prompts that can be used with
ChatGPT at different stages of dissertation writing. Whether you’re
analysing interview data, writing a methodology, or just cleaning up
your references at the last minute, you’ll find something here to make
your life a little easier.

How This Book is Organised

This eBook is divided into sections based on common thesis chapters

and tasks. Each section starts with a short commentary on how
ChatGPT can be useful in that stage, based on my own experience.
Then you’ll find a list of prompts - some direct, some reflective, some
creative - that you can try out yourself, copy-paste, or adapt however
you like.

Customizing ChatGPT before we start

Before you start, it makes sense to customise ChatGPT a little.

2
First, if you want it to write in your style, you’ll need to teach it what
your style is. To do this, paste in a sample of your writing (e.g. a
chapter you’ve written or a previous academic assignment), and say:
“Here is a sample of my writing. From now on, I would like you to
emulate this style when writing for me.”

Also, if you plan to have it write certain sections or chapters, it’s worth
addressing the issue of AI detection. Universities are increasingly
using AI-detection tools (just like they’ve used plagiarism checkers for
years) to identify AI-generated content. At the moment, the situation
is messy - there’s
little consistency
across institutions,
and policies are
vague. But cases do
exist of students
being penalised or
even expelled for
submitting AI-
generated work.

As I said in the
introduction, it’s
important to be reasonable - and I wouldn’t recommend that you have
ChatGPT write full sections for you anyway. But whether you use it a
lot or a little, and even if you later rewrite or restructure the output,
AI detection is a real risk. Worse still, detection tools are often
unreliable and have been known to flag human-written text as AI-
generated.

3
I know this first-hand, as part of my current work involves salvaging
academic writing for students who desperately need to “humanise”
the text they’re about to submit. And (yes, here comes the shameless
self-promo), if you’re in that situation, I do offer that service.

Before it gets to that point, though, here’s a prompt you can use to
significantly reduce AI detection risk. You can insert it after you ask
ChatGPT to write a significant piece of text that you want to include in
your writing:

Write in a way that closely resembles human academic writing and

reduces the likelihood of AI detection. The goal is to produce
academic text that reflects how real scholars write. Include the
following:

Vary sentence length and syntactic structure throughout the text to

avoid rhythmic uniformity

Avoid starting all sentences or paragraphs the same way; use a

range of openings to create natural flow

Use hedging language to express intellectual caution, such as modal

verbs and qualifiers, rather than absolute claims

Integrate theoretical framing and critical perspective, rather than

merely summarising sources or claims

Situate ideas within ongoing scholarly conversations by referencing

debates, tensions, or differing perspectives

Use precise, discipline-appropriate vocabulary that reflects

conceptual clarity and field-specific knowledge

4
Avoid vague generalisations or filler phrases that contribute little to
the argument

Vary the way citations are used and embed them naturally in the
discussion, rather than mechanically attaching them to summary
statements

Avoid repetitive patterns in grammar, punctuation, and connector

words; maintain stylistic and rhetorical variation

Vary sentence length and structure, including the number and types
of clauses. Incorporate both active and passive voice where
appropriate

Ensure the writing demonstrates depth of understanding,

positioning the writer as an active participant in the academic
dialogue

Do not use em dashes or overuse words and expressions that are

commonly identified as AI text giveaway

Now that we’ve addressed these basics and set up ChatGPT for
academic work, let’s explore the prompts you can use at different
stages of your dissertation writing.

5
1. Planning and Idea Development

Before you write anything, you need to figure out what you're actually
researching. ChatGPT won’t do the thinking for you, but it can help
you test some ideas and
explore directions. Here
are some prompts you “Here are my draft research questions.
can use:
Are they clear, specific, and feasible?
What would you improve?”

• I want to
research mental
health support for university students, but that’s too broad.
Can you suggest 3-5 more specific research angles I could
explore?

• Here’s what I’m interested in: [insert paragraph]. Based on

this, can you help me turn it into a feasible dissertation topic?

• I want to explore [here explain your idea and general aim]

Help me turn this into 2-3 research questions that would fit a
qualitative study and would help me achieve this aim.

• Here are my draft research questions. Are they clear, specific,

and feasible? What would you improve?

• Based on these aims, what kinds of research questions would

be realistic and focused enough for a dissertation?

• Here is my research aim: [insert]. Is it narrow and practical

enough for a small-scale MA-level project?

6
 • Can you review this aim and suggest how it could be more
focused?

• My topic is about adult learners returning to education. What

theories might be relevant to this type of study?

• List some commonly used theoretical frameworks in

qualitative research on [topic], and explain briefly what each
focuses on.

• Help me outline a basic timeline for a small qualitative study

that involves interviews and thematic analysis.

2. Research Design

These prompts help you explore methodological options, justify your

decisions, and develop practical research tools.

Here are example prompts, grouped into several categories:

Choosing methodology and identifying philosophical

assumptions

• If I am doing [here explain the procedures/aims/approach to

interviewing, etc], am I doing phenomenology/grounded
theory/ethnography?

• Based on the information about my study so far, briefly

explain what philosophical worldview, epistemological
stance, and ontological assumptions it communicates. Write a

7
 section that not only defines these terms but also provides
arguments, based on my research design and explaining why
these positions are appropriate.

• Can you explain in

simple terms what
grounded theory, “Based on the information about my study so far,
phenomenology, and briefly explain what philosophical worldview,
case study epistemological stance, and ontological
approaches involve? assumptions it communicates”

• Here is my research
topic and question:
[insert]. Which of these approaches would make the most
sense for my study, and why?

• Compare constructivist grounded theory and descriptive

phenomenology in terms of data collection, analysis, and
research focus.

Justifying your research approach

• Based on the research gap, rationale, and my research aims

and questions, help me develop arguments for why [insert
research design, methodology, approach, or method] is
suitable for this study.

• Based on my aims and objectives, explain why qualitative

research is the most suitable approach.

8
 • Based on what you know about my study, write a compelling
argument for why [insert the methodology you have chosen]
is the most suitable choice

• Here are the aims and context of my study. Can you write a
paragraph justifying the use of a case study approach?

• Help me argue that XYZ theory is suitable for the goals and
context of this study.

Sample size

• Is X number of participants sufficient for a [insert

methodology] study exploring [insert research aim]?

• Based on my research question and design, what is a

reasonable range of participants to aim for? Provide
arguments that support your response.

• Compare recent studies using phenomenology in [topic] and

summarise what sample sizes were used to reach saturation.

• Give me a few academic sources (or typical numbers) I can

reference when justifying my sample size in a qualitative
study.

Developing data collection tools

• Based on the [pasted/attached] study aims and literature

review, develop an observation guide, interview protocol, or

9
 focus group plan that will help me collect relevant
information from the perspective of XYZ theory

• Here is information about my study: [insert]. I plan to

interview [describe participants]. Based on this, write a
detailed semi-structured interview guide. Start by writing
down the main questions that correspond to my research
questions. Then, for each, write follow-up questions. Organise
them into sections for the final guide.

• To explore [here insert your statement of the aim], what data

collection methods do you recommend?

• Here is my current draft of an interview guide. Are there any

obvious gaps that would prevent me from answering my
research questions?

• Can you review this interview guide and suggest additional

questions that would help me better address my research
aims?

• I want my interview questions to cover emotional, cognitive,

and behavioural aspects. Can you help me draft prompts for
each?

“Here is my current draft of an interview guide.

Are there any obvious gaps that would prevent
me from answering my research questions”

10
Ethics and consent

• Given that I’m interviewing migrants who may have

experienced trauma, what specific ethical considerations
should I keep in mind?

• How can I phrase a consent form to make sure participants

understand they can stop the interview at any time without
consequences?

• I’m working with a vulnerable population. How can I

minimise emotional distress and still collect meaningful data?
Support your response with references

• Write an informed consent form that includes: an

introduction to the researcher and the study, the selection of
participants, their role and rights, a summary of procedures,
and details about the duration, location of data collection, and
participant expectations.

Pilot testing

• Based on the interview guide, write me a sample interview

transcript which I can use to test my data analysis
procedures.

• Which of these questions might be too leading or too closed for

qualitative data collection?

• Can you help rewrite this question to make it more open-

ended?

11
3. Literature Work and Source Mapping

A strong literature review is the foundation for your study. These

prompts may help you structure your review and identify relevant
theories.

• The aim of my study is … and my research questions are … In

short, my study addresses [describe it], and here is all the
relevant material and notes I want to use in my literature
review chapter. Based on this information, write a chapter
structure outline in which [Topic 1] serves as the key, leading
concept and gradually narrows down toward my study, its
aims, and research questions.

• Help me find as many relevant

Use prompts for
theories and scholarly arguments as
exploring the literature
possible that would support my with caution - ChatGPT
suspicion that […] is not a research
database and may miss
• What theories have discussed the key academic sources.
influence of what we believe about For a more reliable tool
specifically designed for
other people’s perceptions on our
literature exploration, I
self-esteem? recommend SciSpace

• Help me map out 3-5 major themes

commonly explored in the literature on [topic] and explain
how my study fits into or builds on this existing work?

12
 • Write a summary comparing 3–4 theoretical frameworks
commonly used to study [topic]. What are their strengths and
limitations?

• Here’s my draft literature review [insert your literature

review]. Can you help me improve flow and structure by
suggesting where to move or combine sections?

• List 3–5 key authors or studies I should be familiar with when

writing about [theory/topic] and explain why they’re
important.

• Can you write a paragraph explaining why Theory X is

relevant to my study, including references to how it has been
applied in previous research?

4. Coding and Data Analysis

These prompts guide you through initial coding, refining codes,

developing themes, and writing about your findings. (If you’d like a
more detailed breakdown of the full data analysis process, I invite you
to read my e-book on AI-assisted thematic analysis.)

Initial coding

• We will be working on coding my interview transcripts to

conduct qualitative data analysis. I will be pasting
transcripts, one by one. I want you to take each transcript,

13
 and code it with initial codes, otherwise known as open codes.
They should be detailed and descriptive – reflecting what the
passage or sentence is about. I want you to create as many
codes as possible, but do not code the interviewer’s input. As
an output, I want you to create an Excel file with two
columns: left = transcript, right = code.

• Please create the Excel file as requested above, and make sure
it retains the formatting so I can clearly follow which parts of
the text were coded and how.

Focused (axial) coding and grouping

• Now that I have Excel files with initial codes, I want you to
take all codes from the uploaded transcripts and sort them
based on their topic. Create as many groups as needed. If
there are some codes left that don’t fit anywhere, place them
under a group called ‘Other Codes’. Keep original code names
so I can trace them back.

• Now, within each group, check if there are duplicate codes or

near-identical ones. Merge where appropriate, but please list
which original codes were combined into each new one so I
can locate the originals in the Excel files.

14
Developing themes

• My research questions are: [insert]. Based on the transcripts,

the research questions, and the codes you’ve just organised,
come up with themes and sub-themes that will effectively help
me communicate answers to these research questions.

Note: you may prefer to develop themes yourself and only use
ChatGPT to refine them. In that case, once you have a draft of your
themes, you may paste them and ask ChatGPT the following:

• I have developed themes and sub-themes. Please help me

reword them to make them sound more consistent, succinct
and professional.

OPTIONAL ADD-ON (if you want to merge and reorganise your initial
thematic framework): If you see a clearer way to organise them,
suggest a revised structure.

OPTIONAL ADD-ON 2 (if you are at the early stage of theme

development): If you notice repetitive or overlapping sub-themes,
propose merged versions and clearly list which ones were combined.

15
5. Writing and Drafting

These prompts support drafting each chapter and refining how your
findings are communicated.

I’m not a big fan of the idea of having ChatGPT write entire chapters
for me - but as I said at the beginning, I’m not here to judge or tell you
what’s right or wrong. If you do want it to write your chapters for you,
here are some ideas for how to do that (just don’t forget the advice I
shared in the opening sections of this book, where I explain how to set
up ChatGPT to reflect your style and minimise detectable AI content).

Abstracts and summaries

• [Paste your dissertation] Based on this, please write an

abstract that is no longer than 250 words. I want it to
mention the background and research gap, my study aims
and research questions, briefly review the methodology and
key findings, and very briefly mention the main implications
of the findings.

Literature Review
(but see Section 3 for planning and structuring the chapter - these prompts
here are focused on the writing)

• Here is my draft literature review. Can you improve flow and

paragraph transitions without changing the core content?

16
 • I want this section to end with a clear rationale for my study.
Can you help rephrase this paragraph to build that rationale
more strongly?

Methodology

• Here is a paragraph describing my sample and recruitment

strategy. Can you rewrite it more clearly while keeping it
formal and academic?

• Write a short explanation of why semi-structured interviews

are appropriate for this study, based on the following aims
and design: [insert].

Results
(but see Section 4 for coding and theme development)

• I am now writing up my findings. This section focuses on the

theme [insert name], specifically the sub-theme [insert name].
Below are all the relevant quotes for this sub-theme: [insert].
Based on this, write a section that explains what the theme is
and provides a balanced discussion, supported by quotes from
the data.

17
Discussion

• Here is my introduction and literature review: [insert]. Here

are my findings: [insert]. Aims and research questions:
[insert]. Based on this, please write a detailed Discussion
chapter that explains how the findings answer the research
questions and positions the results within the literature
reviewed earlier.

Implications and Contributions

• Attached are the following chapters and sections:

Introduction, Literature Review, Methodology, and Findings.
Based on your understanding of the study’s rationale, the
issues it addresses, and the results, write a detailed
Implications section. If possible, organise the implications into
distinct categories or groups.

• Based on the study’s findings and the gap identified in the

literature review, write a paragraph explaining how this
study contributes to existing knowledge.

Visualising the results

You can either describe your ideas for what a diagram or chart should
include, paste your results and ask ChatGPT to visualise them, or
upload a photo of a rough sketch or doodle and ask it to make it more
professional:

18
 • Based on my sketch [or notes], create a clear diagram that
represents the findings.

• Based on the results, create a clear diagram that shows

thematic relationships in the study.

• Based on the findings, give me ideas for how I can visualize

them.

• I want to show a timeline or process model for how

participants experienced [phenomenon]. Here’s the data - help
me visualise it.

6. Polishing, Referencing, and Formatting

Before you submit, you’ll need to make sure your work is clean,
properly referenced, and consistent. These prompts help you tighten
up your writing:

Proofreading

• Please proofread this and only correct errors but not

structure, so that it is not flagged as AI content

• Here is a paragraph I’m not happy with. Please improve the

grammar and clarity but don’t change the meaning or tone.
Use the same writing style and only correct errors.

19
 • I want to keep the writing in my own voice. Please only flag
clunky or unclear sentences and suggest alternative phrasing
- without rewriting the entire text.

Referencing and citation style

• Here’s my reference list and the [insert the referencing

convention of your choice] referencing guide. Can you check
for consistency and identify anything that doesn’t follow the
correct style?

• Review my in-text citations and make sure they match the

entries in my reference list.

• I’m using APA 7. Based on the guidelines I’ll paste below, can
you check whether the following page of references is
correctly formatted?

Other formatting tasks

• Can you scan this document and identify any acronyms that
need defining or reusing more consistently?

• Please check for any inconsistencies in formatting - such as

use of italics, quotation marks, and subheadings - and suggest
corrections.

• Does this table meet general academic formatting standards?

Let me know if anything looks off or unclear.

20
7. Critical Reflection and Viva Prep

These prompts help simulate difficult questions you may be asked

about your study, and help you prepare responses:

• Based on my study design and methods, what are some

possible weaknesses or limitations I should be ready to
discuss?

• Can you list 5 tough questions an examiner might ask me

during my viva, and suggest how I could respond?

• What are the most likely challenges to using

[methodology/theory] in my study, and how can I justify my
choices?

• Please ask me difficult questions about why I chose this

methodology and these methods. Try to challenge me like a
critical reviewer would.

• If someone asked why I didn’t use a mixed-methods or

quantitative design, how can I respond confidently based on
my aims and questions?

• What would be a clear and persuasive explanation of why I

used semi-structured interviews in this context?

• Make a list of difficult questions I might be asked about my

study during a viva, along with suggested responses.

21
Conclusion

That’s it - no magic, no shortcuts, just a set of tools that can make the
process a little smoother. You don’t need to use every prompt, and you
certainly don’t need to let AI take over your work. But, when used
well, ChatGPT can actually help us become better scholars and
researchers. As always, you’re still the researcher, and you’re still in
control - the AI just helps move things along.

Thank you for trusting me with your time, and your purchase of this
eBook. And thank you for following my work across YouTube, TikTok,
LinkedIn, or wherever else we’ve crossed paths. I still can’t believe
that some of you have been with me since the very beginning - back
when there was no AI on the horizon.

And now here we are, talking about AI detection. I can’t help but
wonder: what will we be dealing with in five more years? Whatever it
is, I hope I’m still helping you make sense of it.

22