Ask Questions First: A Better Way to Design PBL with AI

Apr 28

You need to design a Project-Based Learning (PBL) experience. Maybe it's for an E-Book or a course. You look at your screen. AI tools like ChatGPT seem like the easy answer. So, you ask it directly: "Design a PBL experience for an E-Book about [your topic]."

It feels quick. But the results? Often, they're too general or don't match what you really wanted.

Here’s a simple trick that works much better. Instead of telling the AI what to do, ask it to ask you questions first. Try this: "Ask me 5 questions about designing a project-based learning experience for an E-Book on [your topic]. Pause after each question."

This small change makes a big difference. It's a better way to handle complex design work, like creating educational projects.

Why does asking the AI to ask questions work better? It changes how you work together. When you just give a command, the AI has to guess what you need. It doesn't have enough context about your goals, who you're designing for, or any limits you have.

But asking questions turns the AI into a partner that helps you think, like a guide asking questions.Its questions make you spell out your ideas clearly. Talking it through helps you figure out what makes a good design.

This works especially well for PBL because PBL is complicated. It’s not just about teaching facts. It’s about helping people learn by exploring, working together, and thinking about what they learned. Good PBL design means juggling many things: learning goals, keeping students interested, fair testing, real-world links, and giving students freedom while still guiding them. It needs to be personal and consider the learner's path.1Just asking an AI to "design" all that often skips the important thinking steps. Asking it to ask questions helps you think through the complexity step-by-step.

Many people use direct commands because they seem faster. But for tricky tasks like creating a great PBL experience, that's often not true. When you tell an AI to design something complex, you let it do the hard thinking – figuring out goals, audience, and how to test learning. But the AI doesn't know your specific situation or teaching ideas. It makes guesses, leading to basic results you have to fix anyway. You often spend more time fixing a bad result than you would have spent answering a few questions first. Asking questions first might seem like an extra step, but it actually gets you to a good result faster because you clarify your thinking early on.

Why Asking Questions Helps You Design Better PBL

Just telling an AI "design a PBL" doesn't work well for something complex like teaching design. A simple command like "Write a thank-you email" is fine because the task is clear. But designing PBL is different. It's creative. You need to think about learning goals, who the learners are, the E-Book's style, any limitations, and the teaching choices that make a project great. It's hard to put all that detail into one command. Without that background, the AI gives you something basic.

Asking the AI to ask questions solves these problems in a few ways:

1. It Makes You Think Clearly (Like Socrates Did):

When the AI asks, "What main things should people learn from this PBL E-Book?" or "Who is this for, and what do they already know?", you have to give clear answers. This is like the Socratic method: a conversation using questions to make you think harder, check your assumptions, and understand things better. The AI acts like a curious partner, making you examine your goals, limits, audience, and teaching ideas. Thinking about these things is key to good design. It helps connect your big ideas about education to the actual project structure. It forces you to focus and think about the whole project. Explaining why you want something, because the AI asked, often gives you more insight than just stating it. This back-and-forth supports the kind of exploration and thinking that PBL itself uses.

Sometimes, teachers know their subject well but find it hard to explain the learning steps for a beginner or list every part of a complex project. AI asking structured questions helps with this. The AI doesn't need to be an expert teacher; it just needs to prompt you, the expert, to share what you know. Questions like "How will you check if students understand?" or "What specific skills should students show?" make you state your hidden assumptions clearly. This clarifies the design and can show you gaps or ideas you hadn't questioned before. This leads to a stronger, more learner-focused PBL plan.

2. It Allows Step-by-Step Changes and Structured Thinking:

Unlike a single command, asking questions lets you build the PBL design piece by piece. Your answer to one question helps the AI understand and ask better follow-up questions or suggest ideas. This back-and-forth is like design thinking – you refine ideas and make changes as you go. If an idea doesn't work, you can easily change direction in the conversation without starting over. This is much better than getting a complete but flawed design from a direct command, which often means starting again or doing major, frustrating edits.

This step-by-step process is similar to prompt techniques like Chain-of-Thought (CoT), which break down hard tasks to help the AI think better. Asking the AI to ask you questions does something similar: it guides both you and the AI through the design process in a structured way, making it less overwhelming and more logical. The AI's questions structure the design work itself.

Designing good PBL is complex. It can feel overwhelming, especially if you're new to it or the framework. Asking questions acts like a support system. The AI helps you during the design process. By asking relevant questions (maybe starting broad, then getting specific about activities and testing), the AI breaks the big task into smaller parts. You can focus on one thing at a time. This makes the design process easier to manage, reduces the chance of missing important parts, and leads to a better-thought-out PBL plan.

3. It Helps Personalise and Solve PBL Problems:

Common problems in PBL include creating good tests, adjusting for different learners, managing project details, and balancing student freedom with needed support. Asking the AI questions lets you tackle these problems during the design chat. You can ask the AI (or ask it to ask you) specific questions about these areas: "How will you test teamwork skills?" "What help can you offer students who need more support?" "How will groups manage their time?". This makes sure the final design isn't generic but includes ways to handle common PBL issues.

This conversation also makes the design more personal. The AI learns about your specific situation – the E-Book format, the learners, your resources, and your teaching priorities – from your answers. The final result fits your reality, not just a template. Based on this clear understanding, the AI can help create personalised parts like specific reflection questions, different task options, or even first drafts of grading guides. Studies show AI can help students in PBL feel more capable and connected by offering personalised feedback and adapting to their needs. The interactive design process does this for you, the designer, making you feel more capable as you shape the project with the AI.

Making Your Interactive PBL Prompts Better

Just asking "Ask me questions..." is a good start. You can make it even better by being more specific and combining it with other prompt methods.

Going Beyond Basic Questions:

Instead of just asking for any questions, tell the AI what part of PBL design you want to focus on. Try prompts like:

"Ask me 5 questions about good ways to test learning for this PBL E-Book project."
"Ask questions about problems students might have working together on this project."
"Ask me questions to clarify the final thing students will create for this E-Book PBL."

Giving the AI a role can also change the questions it asks. Try instructions like:

"Act like a sharp instructional designer looking at my first idea. Ask me tough questions about whether this PBL E-Book plan makes sense for learning."
"Act like a curious 8th grader using this E-Book. Ask me what the project is and why it's interesting."
"Use the Socratic method. Ask me questions that challenge the basic ideas of my PBL design."

You can also guide the type of questions, moving from simple clarification to deeper thinking, like Socrates did:

"Ask me questions about the reasons and proof for my project theme."
"Ask me questions about other ways to structure this PBL."
"Ask me questions about the possible effects of the testing method I'm thinking about."

Using This with Other Techniques:

The first question-and-answer part sets a good base. Once you've clarified your PBL E-Book idea through conversation, you can use other prompt techniques better. For example:

Few-Shot Prompting: After talking about testing, give the AI an example of a good grading guide (rubric) for a similar project. Then ask it to create a draft rubric based on the details you've now figured out.
Prompt Chaining: Take the information from your Q&A (like the learning goals and audience) and use it in a new prompt. Ask the AI to suggest specific project activities or create instructions for different learning levels.

Practical Tips:

Using AI this way needs you to participate thoughtfully.

Be Ready to Answer: The AI's help depends on how clear and thoughtful your answers are. Vague answers lead to vague questions.
Guide the Chat: If the AI goes off track, or you want to focus on something specific, tell it directly: "Okay, but let's talk more about X now," or "Can you ask more about the resources students need?".
Stay Critical: Remember, AI can still make mistakes or show biases from its training data. Even though asking questions grounds the AI in your context, always double-check facts or question suggestions that seem off.
Think About Privacy: Be careful with privacy rules, especially when talking about student needs or sensitive info. Don't put personal student data into standard AI tools unless it's allowed and secure.
Expect to Iterate: Think of it as a process. The first round of questions might cover the basics. A second round might focus on improving details.

Asking questions can also help you spot your own biases. When the AI asks why you chose a certain audience or activity, or how you'll make it accessible, you have to explain your thinking. This can reveal assumptions you hadn't examined about fairness or inclusion, leading to a better PBL design.

Working with AI: More Than Just a Tool

Using AI interactively for complex tasks like PBL design, or writing articles like this, changes how we see the technology. It's not just a machine to automate work or give you a finished product quickly. It becomes more like a partner in thinking. AI doesn't replace your thinking, creativity, or teaching knowledge; when used well, it helps and improves it. The conversation becomes a place where you build ideas together. The AI helps structure your thoughts and explore options, while you provide the vision, context, and judgment.

Conclusion: The Power of Asking Questions

Designing good Project-Based Learning is hard. It takes careful thought, teaching insight, and creative problem-solving. AI like Large Language Models can help a lot, but just telling them to create a design often gives you shallow results.

A simple change—asking the AI to ask you questions—leads to a much better design process. This interactive way turns the AI into a partner that helps you think. It pushes you to reflect, allows you to refine ideas step-by-step, and makes sure the final PBL design fits your needs and tackles potential problems. It uses the AI's ability to structure questions while keeping you, the human designer, in charge of the vision and teaching choices.

If you're an educator or creator using AI for complex tasks like making PBL E-Books, starting a conversation is more than just a "trick." It's a strategy that leads to a better process and better learning experiences. Next time you need to design PBL, try asking questions first. See where the conversation takes you.

Phil