This content is brought to you by the Association for Christians in Student Development (ACSD), a volunteer membership organization committed to equipping and challenging faithful professionals to infuse their Christian faith into student development practice and scholarship. Thoughtful content such as this is made possible by volunteer contributions and the financial support of membership dues. Interested in becoming a member for more awesome content just like this? Join today by clicking here!
The use of artificial intelligence in academic spaces is a contentious topic (Sharples, 2023; UNESCO, 2023). I believe the tension can be blamed on the core emotions the topic invokes: excitement, fear, anxiety, and anger foremost among them. As someone frequently in the “Early Adopters” category for new technology, I was among the few who responded to the release of ChatGPT from OpenAI in November 2022 with excitement. As my colleagues suffered from pessimism or dismissal of its relevance, I was scheduled to start my first doctoral-level class a few weeks after the release of ChatGPT. I knew I had an incredibly rare privilege to learn about generative AI as both an educator and a student if I engaged the new tools throughout my coursework.
The initial results were miserable. My first course in January 2023 was Legal Issues in Higher Education. I prompted ChatGPT to help me find legal cases fitting certain criteria that I could use in case studies. If you were using AI models in early 2023, you might predict my issue. ChatGPT seemed to invent from thin air erroneous legal cases that never occurred. In those days, the model was not able to link the sources it referenced. I spent hours writing prompts to ChatGPT, googling the examples it gave me, and returning to reprimand it for falsifying information. Out of frustration, I stopped using the tool in my academic work for nearly two years.
Now, in the 2025-2026 academic year, I am nearing the completion of my doctoral studies. I reengaged AI tools in the summer of 2025 as I worked on my dissertation proposal – fortunately, my program encouraged the creative experimentation and application of AI tools in the dissertation process.
Ground Rule Assumptions
When it comes to technology and learning, I hold some starting assumptions that should be acknowledged. Ultimately, I hope these ground rules provide guardrails, as well as clarity for our engagement with AI as an emerging technology in academia. If we disagree on these points, you will likely disagree with my practical suggestions, but hang with me!
- AI technology is here and cannot be undone. The rate of its growth and its future relevance is debatable; its existence is not.
- A limited baseline of anxiety about the potential nefarious purposes of AI are useful for maintaining ethical boundaries for the tool. Severe anxiety or fear responses are unhelpful and will be largely dismissed by the opportunists who will drive AI.
- AI, like other tools, will enhance the abilities of dissertation writers much like the internet has enhanced thorough literature reviews or like the digital computer simplified typing. IT follows, then, that dissertations published in the age of AI will be held to higher expectations than dissertations of the past, much like how the ability to find every research article ever published online increased the standards of literature reviews of the generation that was limited to hard copies.
- If ethical, competent, high-quality researchers are not the ones to develop best practices for the use of AI in research, that work will be left to the unethical, incompetent, low-quality researchers to develop.
- Though the high energy cost of AI programs is a valid argument to reject or limit the utilization of AI, we must consider that some queries might be worth the cost. Personally, I think research aimed at bettering the lives of students is among those worthy queries.
How to Enhance Your Research with AI
My own engagement with generative AI has been with ChatGPT and Gemini. As such, most of my suggestions are for this or other language model tools, with the exceptions of some use of Google NotebookLM and even less use of ResearchRabbit.
Introduction
The purpose of the introduction is to argue for the relevance, importance, and urgency of your topic. The primary challenge is to produce a clear and defensible logical flow from your opening sentence to your purpose statement. AI can help highlight strengths and weaknesses in your logical reasoning; AI should not be used to create logical reasoning for you.
Dr. Guy White suggested building your introduction chapter on 20 sentences (White, 2015). The first sentence is your hook, the last sentence is your purpose statement. Every sentence in between builds on the last, adds something new to your logical reasoning, and is supported by at least three sources. This exercise creates the logical structure of your introductory chapter. Before you turn the 20 sentences into 20 paragraphs, the final step of this process, input the sentences into ChatGPT to evaluate their logical flow. It is much easier to evaluate the logic of 20 sentences than of a whole chapter.
Literature Review
The purpose of the literature review is to immerse yourself and your readers in what others have said about your topic. The primary challenge is to organize all the existing information into a coherent flow that sets your study in context. AI can help you keep track of all the information you learned while reviewing the literature. Importantly, AI should not be used to generate the information in the literature review.
As you immerse yourself in the literature, keep track of useful direct quotes and their sources. Bonus points if your tracking device is a spreadsheet and you use content tags for sorting! Once you have a critical amount of quotes (400-800 total), upload the document into NotebookLM. This Google tool uses only the uploaded source information to provide questions to your answers. You may consider asking:
- “What have authors suggested for further research?”
- “What major theory areas have been considered for this population?” OR “What populations have already been studied in this theory area?”
- “What is some advice to practitioners based on the research findings?”
NotebookLM can be employed for additional tools, such as an audio overview or a visual presentation. By waiting just 5 minutes for the overview to generate, you’ll be rewarded with a 15-20 minute podcast version of your quotes. As I wrote my literature review, I would listen to this in the shower or while on a walk to keep a summary of all 700 quotes fresh in my mind. This helped with building an outline and with making connections between research studies.
If I were writing my Literature Review now, I would involve a tool called ResearchRabbit. This tool can be greatly helpful in following the thread to identify new articles based on ones you’ve already found.
Methodology
The purpose of the methodology is to describe your study and defend why that method is the best option to answer your research question(s). The primary challenge is to learn about different methodologies, and ultimately choose one that is both manageable for your timeline and strategically the most effective way to answer your research question(s). AI can help you understand how different methodologies would examine your topic. AI should not be used to select a methodology for you, or to make decisions about how that methodology is implemented or analyzed.
One of the biggest ways I found to waste time on a dissertation is to learn about dozens of different research methods that might work. For your methods section, consider using AI as a deeply knowledgeable mentor helping you select the best method for your research problem.
Feed ChatGPT your research problem and describe what you are hoping to learn from your participants. Then, ask how the problem could be addressed through various methods, including quantitative and qualitative aspects. As you narrow in on methods that interest you, ask ChatGPT to compare the differences between those methods. For example, I was debating between phenomenology and interpretative phenomenological analysis (IPA) for my dissertation. By asking ChatGPT about the differences between the two, I realized that IPA was the best choice for my research question and the type of information I wanted to learn. If you still cannot decide, ask ChatGPT to compare the differences in how you would write the research questions based on which method you choose.
Preparing for the Proposal Defense
Upload your program’s dissertation guide to ChatGPT along with your target dates for the proposal defense and final defense. Ask it to provide you a weekly to-do list from now until those target dates of all the things required of you prior to the defenses. Modify that rough draft of a list to what you imagine is doable. If you stick to it, you have a reasonable projection for your graduation timeline.
Once you have a draft of the first three chapters, add ChatGPT as an additional member to your research team with a prompt that reads like this:
Are you ready for a big task? I would like for you to act as a critical peer reviewer for my dissertation proposal on the experiences of gender identity development, spiritual formation, and sense of belonging among transgender and gender nonconforming Christian college alumni. Please evaluate the problem statement’s clarity, the literature review’s depth, the methodology’s alignment with research questions, the feasibility of the timeline, and the potential impact. Also, identify any logical inconsistencies, suggest ways to strengthen the argument, and propose specific questions a committee might ask. Don’t hold back – I want to know how I can improve this draft for my proposal defense. Examine the document thoroughly, I don’t mind if it takes some time for you to generate your response as you apply deep reading, thorough research, and make suggestions that prove you’ve read the document in its entirety.
Whether you agree with its suggestions or not, this might be the first time that someone other than you “reads” your draft in its entirety, and can make suggestions from an unbiased, snapshot perspective.
Summary
AI can be an educational tool. It can be used well or poorly, ethically or irresponsibly in the pursuit of learning and educating. Those of us who are simultaneously students and educators in the early years of AI technology sit in a unique position to experience the tools as a learner and pass on what we discover. The undergraduate students we currently serve will soon step into entry-level employment opportunities that will require AI literacy from them. It is our responsibility to prepare them for that reality. We cannot teach students how to use these tools ethically and effectively if we are not willing to first learn by doing.
So take a little risk, and consider using AI to write your dissertation.
*As a disclaimer, I am an Ed.D. student whose dissertation proposal defense was scheduled around the time I wrote the first draft of this article. All suggestions are for composing your Introduction, Literature Review, and Methodology chapters.
References
- Sharples, M. (2023). Towards social generative AI for education: theory, practices and ethics. Learning: Research and Practice, 9(2), 159–167. https://doi.org/10.1080/23735082.2023.2261131
- UNESCO. (2023). Guidance for generative AI in education and research. UNESCO. https://unesdoc.unesco.org/ark:/48223/pf0000386693
- White, G. E. (2015). Dissertation warrior: The ultimate guide to being the kind of person who finishes a doctoral dissertation or thesis (2nd ed.). Stylus Publishing.
