Session 17: Planning Your Dissertation with ChatGPT
If you are ABD — All But Dissertation — this session is written for you. You have finished the coursework. You have passed your exams. And somewhere between there and the finished manuscript, you have gotten stuck. Today we use one sustained conversation with ChatGPT to move you from a blank page to a real plan: a narrowed topic, a chosen methodology, and a month-by-month timeline you can bring to your chair.
Video Introduction
Watch: Session 17 with Dr. Walter
Meet your instructor and hear directly from someone who defended his own dissertation at 82 — Dr. Walter understands the ABD stall from the inside.
Learning Objectives
What You'll Learn
By the end of this session, you will be able to:
Platform Access
- Use one sustained ChatGPT conversation to move through three linked planning tasks without losing context
- Prompt ChatGPT to narrow a broad topic into a defensible research question using clarifying questions first
- Ask ChatGPT to recommend a research methodology and explain the reasoning behind it in plain language
- Draft a realistic six-month dissertation timeline you can revise with your chair
- Recognize that AI-generated drafts are the beginning of the conversation with your committee — not the end of it
Getting Started with ChatGPT
Follow these steps to access ChatGPT and get ready for today's lesson.
Free Account Required
All platforms used in this course offer free accounts with no credit card required. If you already have an account, simply sign in. The free tier gives you everything you need to complete this session.
Core Lesson
- Go to https://chat.openai.com and sign in.
- Start one new conversation for this session and keep it open the whole time — the power of today's technique comes from staying in a single thread so ChatGPT remembers your topic, methodology, and constraints as you go.
- Have a document open to capture the outputs of each stage — you will paste them into an email or Word document to bring to your chair.
- Bring one broad topic area you have been circling. It doesn't have to be perfectly formed — that's exactly what today is for.
Today's Lesson
Read through this lesson carefully before starting the practice exercises below.
Being ABD is one of the hardest places in graduate school to be, because the structure is gone. There are no more classes, no more assigned readings, no more weekly deadlines from a professor. It is just you, the topic, and the empty page. Many candidates stall here for years. Some never finish. Today we cut through that with a simple, sustained ChatGPT conversation that produces three drafts you can actually take to your chair.
The technique is the point: one conversation, three linked tasks. Because you stay in the same thread, ChatGPT remembers your topic when you ask about methodology, and it remembers your methodology when you ask for a timeline. Each answer builds on the one before it. This is different from opening three separate ChatGPT windows and getting three disconnected answers — and it is the technique we will build on directly in Session 18, The Long Conversation.
Stage one is topic narrowing. Instead of asking ChatGPT to just hand you a research question, you ask it to interview you first. A good opening prompt is: "I am an ABD doctoral candidate interested in [broad topic]. Before you suggest anything, please ask me three clarifying questions about my interests, my target participants, and whether I lean qualitative, quantitative, or mixed methods. After I answer, propose three specific, defensible research questions I could pursue." The clarifying questions matter — they force you to say out loud what you actually care about, which is often the real bottleneck.
Stage two is methodology. In the same conversation, you ask: "Based on the research question we just settled on, recommend a research methodology and explain your reasoning in plain language. Then describe what data collection would concretely look like — sample size, recruitment, interview or survey questions, and how long it would realistically take." Because you are in the same thread, ChatGPT ties the methodology directly to the topic you narrowed in stage one.
Stage three is the timeline. You ask: "Now build me a realistic six-month dissertation timeline based on the topic and methodology we have discussed. Include monthly milestones for IRB approval, recruiting participants, data collection, analysis, writing chapters four and five, committee checkpoints, and a final revision month before defense." What comes back will not be perfect. It will be a first draft — and that is the whole point. First drafts are what unstuck candidates.
Here is the critical framing: what ChatGPT produces in this session is a starting document, not a finished plan. Take it to your chair. Say, "I drafted this to give us something concrete to react to — where am I off?" That conversation with your chair, informed by a real draft, will get you further in one meeting than three months of staring at the wall. AI can carry you eighty percent of the way there in about thirty minutes. Your chair, your committee, and your own scholarly judgment carry you the last twenty.
Hands-On Practice
Practice Exercise
Do this in one continuous ChatGPT conversation. Do not open a new chat between steps — the whole point is the sustained thread.
Try These
- Open one new ChatGPT conversation. In it, type: "I am an ABD doctoral candidate interested in [YOUR BROAD TOPIC]. Before you suggest anything, please ask me three clarifying questions about my interests, my target participants, and whether I lean qualitative, quantitative, or mixed methods. After I answer, propose three specific, defensible research questions I could pursue." Answer the clarifying questions honestly.
- Pick one of the three research questions ChatGPT proposes — or ask it to combine two — and confirm your choice back to it in the same conversation.
- Still in the same conversation, ask: "Based on the research question we just settled on, recommend a research methodology and explain your reasoning in plain language. Then describe what data collection would concretely look like — sample size, recruitment, interview or survey questions, and how long it would realistically take."
- Still in the same conversation, ask: "Now build me a realistic six-month dissertation timeline based on the topic and methodology we have discussed. Include monthly milestones for IRB approval, recruiting participants, data collection, analysis, writing chapters four and five, committee checkpoints, and a final revision month before defense."
- Copy the three outputs — narrowed question, methodology plan, six-month timeline — into a single Word document. This is your draft plan.
- Email your chair and ask for a thirty-minute meeting to react to the draft. Bring the document. Do not present it as a final plan — present it as "something concrete to react to."
Example Prompts to Try
These are the exact prompts used in the lesson, formatted so you can copy them directly into ChatGPT. Replace the bracketed placeholder with your own topic.
Summary
Key Takeaways
Dissertation Planning in One Sustained Conversation
You have used a single ChatGPT thread to move from a broad topic to a narrowed research question, from a research question to a matching methodology, and from a methodology to a realistic timeline. That is not a shortcut around your committee — it is a way to arrive at your next chair meeting with something real on the table.
- Being ABD is a stall problem, not an intelligence problem — you already have what it takes; what you need is momentum.
- One sustained ChatGPT conversation is more powerful than three separate ones, because each stage remembers the one before it.
- Ask ChatGPT to interview you before it proposes a research question — the clarifying questions are where the real thinking happens.
- Treat AI-generated plans as first drafts to react to with your chair, not as finished products.
- If you are ABD and stuck, dedicated coaching can shorten the last mile — see our ABD Coaching program for personal, one-on-one support from someone who defended his own dissertation at 82.