Turning Mistakes into Moments of Learning

Every classroom has them—the confident slip-ups, the small errors that make us smile. But what if those mistakes weren’t just corrected and forgotten? What if they could become the most powerful teaching tool you have?

In this episode, we dive into a brand-new systematic review (Dieterich, Rumann & Rodemer, 2025) on learning from erroneous examples. The research shows that when teachers strategically use mistakes—by highlighting them, contrasting them with correct solutions, and giving the right kind of feedback—students don’t just remember the right answer. They build deeper understanding, stronger long-term memory, and the courage to take risks in their learning.

We unpack:

• ✅ Why contrasting erroneous examples (wrong vs. right side by side) outperform just showing the correct method.

• 🧠 How to manage cognitive load so mistakes help rather than overwhelm.

• ✏️ Practical classroom strategies—like circling errors in red, tailoring prompts to student ability, and revisiting common mistakes—that work even in large Bangladeshi classrooms with only chalk and a board

• 🌱 How embracing mistakes shifts classroom culture, helping students overcome the fear of failure and see errors as a normal, valuable part of learning.

Whether you’re teaching 20 students or 60, this episode gives you research-backed, practical ways to turn errors into moments of real learning.

👉 Pair this episode with our blog Turning Mistakes into Moments of Learning  for classroom-ready examples and strategies.

🔑 Key Takeaways

• Mistakes are powerful learning tools — when used intentionally, they help students build deeper, longer-lasting understanding.

• Contrasting erroneous examples (wrong vs. right side by side) are the most reliable strategy, especially for students with lower prior knowledge.

• Erroneous examples on their own can be highly effective, but they require scaffolding and careful prompts to avoid overload.

• Prompts matter — tailor your questions to students’ level: challenge confident learners with “find and explain,” while guiding struggling learners with “look and compare.”

• Feedback should explain the “why” — quick process-oriented comments beat a simple ✓ or ✗.

• Revisiting common mistakes in the following lesson reinforces memory and helps normalise errors as part of the learning journey.

• Classroom culture shifts when mistakes are welcomed — turning errors into teachable moments reduces the fear of failure and encourages curiosity.

Dieterich, S., Rumann, S., & Rodemer, M. (2025). Conditions for effective learning from erroneous examples: A systematic review. Educational Psychology Review, 37, Article 94. https://doi.org/10.1007/s10648-025-10071-x

Podcast Transcript: Stop Hunting! The Cognitive Science of Learning from Mistakes
Speaker 1: Welcome to the deep dive, uh, the research bite series where we give you small bites of big evidence. Today we’re digging into a really interesting systematic review from 2025. It’s all about learning from mistakes in the classroom. The title is uh conditions for effective learning from erroneous examples.
Speaker 2: That’s right.
Speaker 1: And for you, our listener, so especially thinking about teachers in places like Bangladesh, maybe with larger class sizes, maybe fewer digital tools. Well, this deep dive has a really specific mission. We’re not just asking our takes useful. We want to find those practical evidence-based ways you can actually use errors, use them strategically to really boost student learning. You know, using the tools you already have, chalk, board, maybe paper, and just smart instructional design.
Speaker 2: Absolutely. And to get started, we should probably talk about examplebased learning or EBL. This is basically where you show students a problem being solved step by step. It’s a really common technique.
Speaker 1: right? Like a worked example.
Speaker 2: Exactly. And the big reason it works so well, especially for learners who are new to topic is that it cuts down on what we call extraneous cognitive load.
Speaker 1: Cognitive load. That’s like the brain’s working memory capacity, right? How much stuff it can handle at once.
Speaker 2: Precisely. And extraneous load is the unhelpful kind. It’s the mental effort wasted just trying to figure out confusing instructions or searching randomly for an answer. EBL helps focus that limited mental energy purely on the learning itself, on building those knowledge structures or schematada.
Speaker 1: Okay, so this review looked at two specific ways to bring errors into that EBL process. What were they?
Speaker 2: Well, the first one is pretty straightforward. Erroneous examples or we can call it erics. This is simply a worked out problem, but it has a mistake baked into it deliberately and usually it’s presented as if say a fictional student made the error. This helps normalize mistakes, makes it less threatening.
Speaker 1: Okay, so that’s Eric’s just the mistake. What was the second one?
Speaker 2: The second one and this turned out to be really key in the research is the contrasting erroneous example or coner. contrasting.
Speaker 1: Yeah.
Speaker 2: So with conerics, you show the incorrect solution, the erics, right next to the fully correct solution side by side.
Speaker 1: Ah, okay. So you’re explicitly highlighting the difference, the flaw, and the fix.
Speaker 2: Exactly. It’s a direct comparison. It forces the student to see precisely where things went wrong and what should have happened instead.
Speaker 1: It sounds like a powerful teaching tool, that direct contrast.
Speaker 2: It really is. And the theory behind why this works, uh, it goes back to something called the theory of negative knowledge.
Speaker 1: Negative knowledge sounds intriguing.
Speaker 2: It is. The basic idea from Ozer and his colleagues back in 99 is that learning gets a major boost when students have to actively engage with what’s incorrect. This negative knowledge understanding why something is wrong actually strengthens their understanding of what is correct. It defines the boundaries. You see.
Speaker 1: so mistakes aren’t just bumps in the road, they’re signposts.
Speaker 2: That’s a great way to put it. They become really valuable raw material for deeper understanding.
Speaker 1: Okay, that makes a lot of sense intuitively. But let’s get to the core results. Does the evidence actually show these error-based examples work better than just showing the correct way, the standard worked example or correex? Right, the big question.
Speaker 2: Well, the answer is clearest when we look at that contrasting format, the Coner X. When researchers compared Conex against just showing the correct example, Conerex came out on top in nine out of 12 comparisons.
Speaker 1: Nine out of 12. That’s pretty significant.
Speaker 2: really is, especially in subjects like mathematics, procedural tasks. The message there is quite strong. If you can set up that sideby-side comparison, maybe on the board or a handout, the research suggests it’s likely worth the effort.
Speaker 1: Why do you think that contrast is so effective?
Speaker 2: Well, it seems to directly tackle misconceptions. If you only ever see the right way, you might not even notice the common traps or pitfalls. But when you force that comparison, wrong way here, right way there, students build not just positive knowledge like how to do it, but also that crucial negative knowledge how not to do it. It stops misconceptions from forming or helps clear up existing ones.
Speaker 1: Okay. So, Conerex looks like a winner. What about just the mistake on its own? The pure Erics format. You said the results there were different.
Speaker 2: Yeah, they were definitely more mixed. When Eric was compared against the standard correct example, uh nine comparisons did show a benefit for Eric, but importantly, 11 comparisons showed no significant difference. And there was even one study where the correct example actually worked better.
Speaker 1: So that sounds a bit more well high stakes maybe
Speaker 2: right
Speaker 1: if just showing the error might not help or could even be worse in rare cases is it risky for a teacher with limited time is coneric just the safer bet overall
Speaker 2: that’s a very fair point is certainly the lower risk option especially and we’ll get into this for students who don’t have a strong background in the topic yet okay but and this is why Eric alone is still really interesting when it did work better than the correct example it often outperformed it significantly in specific areas
Speaker 1: like what
Speaker 2: primarily in developing that negative knowledge the students ability to actually spot errors and explain why they’re wrong and also in improving procedural knowledge the howto
Speaker 1: ah so it forces a deeper kind of thinking when they have to find the error themselves
Speaker 2: it seems so that extra mental effort involved in detecting the flaw diagnosing it and figuring out the correction seems to lead to a more robust critical understanding for some learners and wasn’t there something interesting about when those benefits showed up, the timing.
Speaker 1: Yes, absolutely crucial point. When researchers compared both error types, Eric and Conerex against just traditional problem sol, you know, just just giving students problems to solve from scratch, right? Both Erics and Coner often did better. But here’s the nuance. For Eric especially, those benefits sometimes only really became clear on delayed post tests.
Speaker 1: Oh, delayed like weeks later.
Speaker 2: Exactly. Not immediately after the lesson, but maybe a few weeks on the line.
Speaker 1: Wow. What does that suggest is happening?
Speaker 2: It points towards deeper processing. It suggests that wrestling with that mistake without the immediate answer provided forces the student to really dig into their memory, retrieve the rule, analyze the error, generate the fix. It’s a more demanding process.
Speaker 1: So, it sticks better long term.
Speaker 2: That’s the idea. It seems to promote better schema consolidation, really cementing that understanding into long-term memory, which is obviously a huge plus if you’re aiming for lasting learning, not just cramming for a test.
Speaker 1: Okay, so that sets the stage really well. Errors can be powerful. Conerics is reliable, maybe the go-to starting point. Eric is potentially deeper, but maybe trickier to use effectively. But the review really hammered home that the conditions are critical. So for our listeners, maybe teaching 50, 60 kids relying on chalk and talk account, what are the practical things they can actually control to make this work? The active ingredients.
Speaker 2: Great question. The first big one is prompts, the instructions you give the students.
Speaker 1: like find the error or explain why this is wrong.
Speaker 2: Exactly. Things like explain, reflect, identify, correct, compare. These prompts are what actually drive the student to engage mentally with the mistake. You could have the best error example in the world, but if the prompt isn’t right, the learning opportunity might be missed.
Speaker 1: But you mentioned earlier, it’s not one sizefits-all with prompts, is it? It depends on the student’s background knowledge.
Speaker 2: Precisely. This is critical. The cognitive demand of the prompt, how much thinking it requires, absolutely must match the learner’s prior knowledge, their PK.
Speaker 1: Oh yeah. If you give a really demanding prompt to a student who’s struggling with the basics, you just overwhelm their working memory. You cause cognitive overload and learning shuts down. So how does that play out practically? Say in that large classroom in Bangladesh using just a whiteboard.
Speaker 2: Okay, let’s picture it. For your students with high prior knowledge (HPK), the ones who already have a decent grasp, you can use those more demanding prompts. You could write anx, okay, find the mistake in step three and explain why it breaks the rule we learned yesterday. That forces detection and explanation. Great for them.
Speaker 1: But for the students who are maybe less confident, the low PK students (LPK), that would be too much.
Speaker 2: Way too much. They’d spend all their mental energy just trying to find the mistake, let alone understand or fix it. That hunting is pure extraneous load. It’s wasted effort.
Speaker 1: So, what’s the strategy for them?
Speaker 2: Load reduction. You have to make it easier for them to focus on the concept, not the search. The single most effective thing, the teacher needs to explicitly highlight the error.
Speaker 1: Ah, so physically point it out.
Speaker 2: Yes. Use that bright red chalk or circle it clearly on the board or handout. Remove the where’s Wall-E aspect of finding the mistake.
Speaker 1: Okay, makes sense. Error located. Then what prompt do you use?
Speaker 2: A much simpler, more targeted prompt. Something like look at the step I’ve circled in red. Can you explain why that number is wrong? Or compare this circle step to the correct example we just did. What’s the difference?
Speaker 1: Got it. So, you reduce the search load, then use a focus prompt on the specific error itself, maximizing the useful thinking, the germaine load.
Speaker 2: Exactly. You’re guiding their attention right to the point of difficulty and asking them to process that rather than getting lost searching. That’s incredibly practical. Okay. Highlighting the error for LPK students. What was the second key factor from the review? Feedback.
Speaker 1: Yes, feedback. We know just saying wrong isn’t very helpful, right? Right. The review talked about different kinds, didn’t it? Like just knowing if you’re right or wrong versus getting the actual answer versus getting an explanation.
Speaker 2: That’s right. They distinguish knowledge of result (KR) just right wrong. Knowledge of correct response (KCR) here is the right answer. And crucially, elaborated feedback (EFB).
Speaker 1: Elaborated feedback. So the why?
Speaker 2: exactly. EFB is feedback that explains why the error is an error. It gives hints or explains the process that went wrong. and the review strongly suggests this is critical for building that negative knowledge and really learning from the mistake. Students need to understand the breakdown in reasoning.
Speaker 1: So for the busy teacher marking work or maybe going over an example on the board, what does that mean in practice?
Speaker 2: It means prioritizing brief, specific processoriented explanations over just a ticker across. If you’re showing a common error on the board, instead of just putting a big X through it, maybe quickly say or write, “Careful here. The negative sign wasn’t distributed to both terms. terms inside the bracket. It’s a small shift, but it moves from just judging the answer to providing immediate instructional value about the process.
Speaker 1: That makes sense. Connect the feedback directly to the thinking error, which brings us back, I suppose, to that central challenge you mentioned, prior knowledge and managing cognitive load. It feels like this is the tightroppe teachers have to walk with error examples.
Speaker 2: It really is the crux of it, especially for the pure ERIC format. The research is clear asking students to find and fix an error does increase cognitive load compared to just studying a correct example. They have to detect, diagnose, and correct. That’s a demanding sequence.
Speaker 1: So teachers have to be really mindful of that load, especially for learners who are still building their foundation.
Speaker 2: Absolutely. And this is exactly why Conerex the contrasting format is so effective particularly for those low prior knowledge LPK students.
Speaker 1: because the correct answer is right there.
Speaker 2: Precisely. By providing the correct solution alongside the flawed one, the student does doesn’t have to dredge up the correct procedure from memory and figure out the error. They just need to compare the two.
Speaker 1: Ah, so it reduces the number of things they have to juggle mentally.
Speaker 2: Exactly. It simplifies what we call element interactivity. How many pieces of information they need to hold and manipulate in their working memory at once. The comparison makes the cognitive load much more manageable for LPK students.
Speaker 1: Okay, so let’s say a teacher does want to try the pure ERX format, maybe aiming for that deeper processing even with some LPK students. What are the absolute mustd dos to manage that load?
Speaker 2: Two non-negotiables based on this review. First, explicitly highlight or mark the error. Remove the search task. Second, use simple, highly focused prompts.
Speaker 1: Got it?
Speaker 2: The goal always has to be maximizing the germaine load, the mental effort spent understanding the actual principle or procedure while ruthlessly minimizing the extraneous load caused by confusing presentation or asking them to do too many things at once. Keep the instructional design clean, clear, and focused right on the error. itself.
Speaker 1: This has been brilliant. It feels like we’ve unpacked a really powerful strategy that doesn’t cost anything except thoughtful planning. So, to summarize, errors are definitely learning opportunities. That sideby-side comparison, Coner X seems like the safest, most reliable starting point, especially if you’re unsure about students prior knowledge.
Speaker 2: Right? It manages the load effectively.
Speaker 1: But the real key, whether using conerics or the potentially deeper ERX, is the instructional design around it. Those props and the feedback are what make it work. And critically, as you said, this isn’t about fancy tech. It’s about smart pedagogy you can implement with basic resources. Highlighting errors physically, using specific verbal prompts, giving brief explanatory feedback, these are all doable. And the core principle for helping struggling learners is always reduce that search load. Point them directly towards the cognitive conflict you want them to resolve.
Speaker 2: So maybe the question for you listening as you plan your lessons is, am I creating opportunities for students to learn from errors and am I giving them the right tools like those clear comparisons, the highlighted errors, the focused prompts to actually do that effectively? Am I giving them permission and scaffolding to learn from mistakes? And maybe one final thought to chew on building on that delayed post- test finding.
Speaker 1: Yeah, H like revisiting.
Speaker 2: Yeah, perhaps dedicating just a few minutes to explicitly revisit the reason behind a common error from the previous day. Could that help further stabilize that long-term negative and positive knowledge? It suggests the power of the error can be strategically reinforced over time beyond just the initial lesson. Something to think about.