Lecture slides are a map, not the territory. They compress a topic into bullets, diagrams and definitions, and it is easy to leave a lecture feeling informed while being unable to explain the material aloud. A teach-back routine fixes that gap: you extract the core ideas, prepare a short explanation in your own words, then test yourself with questions a curious classmate might ask. The goal is not a polished performance. It is to expose the places where your understanding is thin.
Step 1: Reduce the deck to five concepts
Open the slides and read them once without highlighting anything. Then write down the five ideas the lecture would collapse without. For a water cycle lecture, that might be evaporation, condensation, precipitation, collection and transpiration. If a deck yields more than five, group related points under a broader heading rather than memorising every bullet.
Next to each concept, add two things: a source location (slide number, section heading or lecture timestamp) and a one-line definition in your own words. This is where AI can help with the mechanical part. Oyriva can work with uploaded supported slide materials and surface cited source passages when available, so you can check a definition against the slide it came from. File support and AI usage depend on your plan and settings, and not every format or layout is preserved, so keep the original deck open as the reference copy.
Step 2: Write a 90-second explanation
Set a timer and explain the topic aloud as if teaching someone who missed the lecture. Speak from your concept list, not the slides. Record it or dictate a draft if that helps you notice gaps. Oyriva's dictation appends speech to an editable unsent draft, and Live transcribes speech into regular chat and speaks responses, which can make a spoken rehearsal easier to review.
Precision matters more than fluency. In the water cycle example, many students say evaporation and condensation almost interchangeably. A teach-back forces the distinction: evaporation is liquid water becoming vapour, usually driven by heat; condensation is that vapour becoming liquid droplets, as when moist air cools. If your explanation blurs the two, you have found a real gap rather than a memory slip.
Step 3: Ask probing questions
After the explanation, write five questions that go beyond definitions. Good prompts include: What causes this step to happen? What would change if one variable increased? Where does this concept connect to last week's lecture? What is a common misconception? For the water cycle, ask how particles in the air can help cloud droplets form, or what happens to precipitation over land versus ocean.
Answer each question without looking at your notes, then check against the slides and any cited passages. If you use an AI tutor to generate practice questions, treat its output as a prompt for your own thinking, not as an answer key. Generated questions and presentations can be wrong, so verify them against the source material.
Step 4: Correct, then retrieve again
Mark every place your explanation drifted from the source. Rewrite only those lines, then close everything and repeat the 90-second explanation from memory. Spacing this over two or three days is more useful than repeating it five times in one sitting. A short retrieval quiz, built from your five concepts, gives you a quick check before the next lecture.
Keep the distinction between making materials and learning them. Oyriva can generate presentations, but a generated deck is not proof that you understand the topic. Presentation generation is a production task; teach-back is evidence of learning. Reading and reviewing are also not graded evidence of mastery. If you submit AI-assisted work, disclose it where your course requires it, and follow your institution's rules.
A weekly routine you can actually keep
After each lecture, spend ten minutes reducing the deck to five concepts with sources and definitions. The next day, record a 90-second explanation and list three probing questions. On day three, answer the questions from memory, correct your notes, and run a short retrieval quiz. Before the exam, revisit only the concepts you explained poorly. If your slides arrive as PDFs, the workflow in turning lecture PDFs into practice questions pairs well with this routine, and verifying AI study answers against sources helps you keep the correction step honest.
Conclusion
Explaining a topic is a diagnostic, not a performance. Five concepts, a short spoken explanation, probing questions and a correction pass will tell you more about your understanding than another read-through of the slides. Keep the original deck nearby, verify anything an AI tool produces, and let the gaps you find guide what you study next.
Ready to try it? Create an Oyriva account to upload supported materials and rehearse your explanations, or compare plans to see which file and AI usage limits fit your study week.
