Here's something that's well established in cognitive science but rarely applied in online courses: the act of trying to recall information strengthens memory far more than re-reading it. Roediger and Butler's research on the "testing effect" (2011) found that retrieval practice produced 10-20% better retention than re-studying the same material. Your students don't need another PDF summary — they need activities that make them retrieve and apply what they just learned. ChatGPT can generate those activities from your lesson content in minutes.
What you’ll walk away with:
- End-of-module review activities reinforcing learning
- A mix of recall, application, and reflection questions
- A reusable template for every module
Why most courses skip review activities
I've looked at thousands of courses on Ruzuku over the past 14 years, and the pattern is consistent: creators pour energy into the lessons themselves — the content, the videos, the resources — and then move straight to the next module. Review activities, when they exist, are usually an afterthought. A quick "reflection journal" prompt. A quiz with obvious answers.
The reason is understandable: creating good review materials is time-consuming. Summarizing a module, writing thoughtful reflection questions, designing an application exercise — that's another hour of work per module, and most independent creators are already stretched thin. AI doesn't make this work unnecessary, but it makes it feasible.
The payoff is worth it. Karpicke and Blunt (2011, Science) found that students who practiced retrieval retained 50% more than students who created concept maps — and concept maps are already considered an active learning strategy. Simple recall questions after a lesson aren't just "nice to have." They're one of the most evidence-backed learning interventions available.
Three types of review activities that work
1. Key concepts summary
Not a copy of your lesson — a distilled version that highlights the 3-5 most important ideas. Better yet, give students the summary with deliberate gaps and ask them to fill in the missing concepts from memory. This "fill in the blanks" approach forces retrieval in a low-stakes way.
2. Reflection questions
Questions that ask students to connect the material to their own work or experience. "Think about a student you've worked with who struggled with [topic]. How would you apply [concept from this module] differently?" These aren't comprehension checks — they're integration exercises that help students see how the material fits into their existing knowledge.
3. Application exercises
A concrete task that requires using the module's concepts. If you're teaching coaching techniques, the exercise might be: "Write a 3-sentence response to this client scenario using the framework from this module." Application exercises work best when they mirror what students will actually do in their professional practice.
Prompts to try
Comprehensive module review
"Here's the content from my course module on [paste lesson content]. Create an end-of-module review activity with three parts: (1) A 'key takeaways' summary — 5 bullet points covering the most important concepts, written in student-friendly language. (2) Three reflection questions that ask students to connect the material to their own work (not recall questions — application and synthesis). (3) One practical exercise where students apply a specific concept from the module to a realistic scenario. The whole review should take a student 10-15 minutes to complete."
Fill-in-the-blanks retrieval exercise
"Based on this lesson content [paste text], create a 'test yourself' exercise with 8 fill-in-the-blank statements. Each blank should require the student to recall a key term or concept from the lesson — not trivial details, but the core ideas they need to remember. Include an answer key at the bottom. Keep the language conversational."
Scenario-based application quiz
"I teach [subject] to [audience]. This module covered [topic]. Write 4 scenario-based questions where each scenario describes a realistic situation my students might face. For each scenario, give 3 possible responses and explain which response best applies the concepts from this module and why. This isn't a trick quiz — it should reinforce the learning, not test for gotchas."
The human layer
AI-generated review activities have a predictable weakness: they're generic. They cover the right concepts but miss the nuance of your specific teaching context. Here's where your expertise matters most.
You know which concepts your students consistently struggle with — those need extra review attention, not the concepts that are straightforward. You know the common misconceptions in your field — reflection questions should surface those, not just confirm what students already understand. And you know the real-world scenarios your students actually face — application exercises should use those, not textbook hypotheticals.
My approach: generate the review activities with AI, then spend five minutes editing each one. Replace the generic scenario with one from your actual experience. Add a reflection question about the mistake you see students make most often. Swap in vocabulary your students actually use. Five minutes of editing turns AI-generated content into something that feels like it was written by someone who's been in the room with your students.
What it gets wrong
- Defaults to recall over application. ChatGPT gravitates toward "What are the three types of..." questions. You'll need to explicitly ask for application and synthesis, and you'll often need to push back when it reverts to quizzing.
- Summaries can be too complete. A summary that covers everything defeats the purpose — it lets students recognize rather than recall. Edit summaries to be slightly incomplete, nudging students to fill gaps from memory.
- Scenarios feel hypothetical. AI creates plausible-sounding scenarios, but they often lack the messy specificity of real life. "A client asks for advice on..." doesn't have the same weight as a story from your own practice. Swap in real examples wherever possible.
- No adaptive difficulty. AI generates one-size-fits-all review activities. In reality, some students need the basics reinforced while others need challenge questions. If your course serves a wide skill range, consider generating two difficulty levels and letting students choose.
Related guides
- Building Self-Grading Quizzes with Google Forms — automate scoring for knowledge checks
- Creating Course Flashcards with Quizlet — spaced repetition for key concepts
- Generating Discussion Prompts with AI — extend review into group conversation
- Creating Gamified Quizzes with Kahoot — live review sessions for cohort courses
Make review part of the learning
The research is clear: students who practice retrieval learn more than students who re-study. AI makes it practical to add review activities to every module, not just the ones where you had time. Start free on Ruzuku and build courses where every module ends with activities that help your students actually retain what they've learned.