BrAInstorming for Research
💡 Core Idea: AI is not always reliable at giving answers, but it can help us ask better questions.
In this learning module, you will learn how to:
- Move from vague topics to genuine research questions.
- Notice how prompts carry assumptions.
You may have already used an AI tool to brainstorm ideas for a paper or project. Let’s take your experience to the next level by incorporating iterative prompts.
A typical brainstorming prompt may be something like,
Give me some topics for a research paper.
This is just a starting prompt! What you really need is a research question. Keep going!
One of the greatest strength of generative AI is that it can have a conversation with you. Conversation involves multiple rounds of back-and-forth sharing rather than just one question and one answer. How long and how deep can you go?
Try out follow-up prompts like,
- What questions might a historian ask about this topic?
- What controversies exist in this topic?
- What are common questions scholars or writers explore related to this topic?
- What parts of this topic tend to generate the most discussion or disagreement?
🎓 PRO TIP: You can ask the AI tool to interview you about a topic you are interested in. It might help you figure out why you care, what you know, and (most importantly) what you don’t know.
Try this prompt:
Interview me about my interest in [topic]. Ask one question at a time. Ask questions that invite me to recognize what matters and what I’m unclear on.
Perhaps your AI conversation yields a topic that intrigues you, or maybe you started off with a topic already in mind. Turn that general curiosity into something researchable. Here are some ideas of prompts you could use to help you brainstorm a research question:
- I’m interested in [topic]. What are some meaningful research questions people ask about this?
- What are different ways people might approach or interpret this topic
- What perspectives are often included—and which are overlooked?
- How might different communities or disciplines view this issue differently?
- What makes this topic complicated rather than simple?
- This topic feels too broad. What are some ways to narrow it?
- What specific aspects of this topic could I focus on?
- What time periods, locations, or groups are often studied in relation to this?
Keep asking questions. Challenge the answers that the AI tool gives you.
❗WARNING: Generative AI will always give you an answer, even if your question contains erroneous assumptions. Reflect on your own prompting process. What are you assuming to be true? What prompts would you give the AI tool if that assumption was not true?
For example,
Give me an unbiased overview of this topic.
This prompt assumes that being unbiased is achievable. In reality, when it comes to information, nothing is truly unbiased. Even an AI overview or summary will reflect choices of what to include, what to emphasize, and what to omit. The AI tool will probably default to the dominant perspective on the topic—which isn’t unbiased at all.
If you land on a question that seems promising, keep going!
Experiment with prompts like,
- Can you help me revise this question to make it more focused?
- What makes this question unclear or too vague?
- How could I reframe this question to invite exploration rather than a yes/no answer?
Revising your question is a valuable research practice.