Learning From Mistakes: How to Use an Error Log to Study

Learning from mistakes works when you do three things: see the correct answer soon after the error, figure out why you got it wrong, and retest the same idea a few days later. Done that way, getting something wrong can teach you more than getting it right the first time. Done badly, you just repeat the error on the exam.

Most students look at a returned test, wince at the score, and file it away. That throws out the most useful study material they'll ever get: a precise list of what they don't know yet.

This guide covers the research on errorful learning, the hypercorrection effect, and productive failure, then gives you an error log template and a weekly routine for using it before exams.

Why does learning from mistakes work?

Because producing an answer, even a wrong one, seems to prepare you to learn the correct one when you see it. In a 2017 review of the research, Janet Metcalfe concluded that making errors and then getting corrective feedback helps learning, and that feedback which includes the reasoning behind the mistake is crucial.

Three findings show what that means in practice:

• Correct answers after errors matter enormously. In a 2005 study, 258 people learned Luganda-English word pairs. Being shown the correct answer after a wrong response raised what they remembered on a test a week later by 494%, roughly six times the no-feedback level. Feedback after correct answers made little difference.
• Related errors help more than random ones. When word pairs were meaningfully related, guessing first and then seeing the answer beat simply studying the pair. With unrelated pairs, guessing gave no benefit. And participants didn't realize guessing had helped them, even right after the task.
• Guessing before you study works too. Trying to answer questions before reading the material, then checking, improves learning even when most first guesses are wrong. This is the basis of pre-testing.

The catch is that errors only help when they're corrected. A wrong answer you never check is just practice at being wrong. That's why every form of active recall needs a checking step.

What is the hypercorrection effect?

The hypercorrection effect is the finding that mistakes you were confident about are more likely to be corrected than mistakes you were unsure of. Brady Butterfield and Janet Metcalfe found this in 2001 using general-knowledge questions: people rated their confidence, got feedback, and were retested. The errors they had been most sure of were the ones they fixed most often.

One likely reason is surprise: being confidently wrong grabs your attention, so you study the correction more closely. Another is that confident errors tend to come from topics you partly know, so the correct answer has something to attach to.

It isn't just a lab effect. In a 2018 study in an introductory college horticulture course, students answered questions on course content, rated their confidence, saw the correct answers, and later took a posttest. High-confidence errors were corrected more often, and students with more prior knowledge were both more confident in their errors and better at correcting them.

Two caveats change how you should use it:

• It may not last on its own. In a 2023 classroom study by Corral and Carpenter, the effect showed up when students retook the same questions but did not carry over to an end-of-course exam with different questions. Retest confident errors later, in a new form.
• Lucky guesses need attention too. A 2008 study found that feedback doubled retention of answers people got right but weren't sure of. A correct guess is not knowledge. Log it like a mistake.

What is productive failure?

Productive failure means trying to solve a problem before you've been taught how, then getting the instruction. You'll usually fail to find the standard method, but the attempt shows you what the method is for.

A 2021 meta-analysis by Tanmay Sinha and Manu Kapur pooled 53 studies with 166 comparisons. Problem solving followed by instruction beat instruction followed by problem solving, with a moderate effect (g = 0.36). The effect was stronger when the approach was designed carefully, and it reversed for young children (second to fifth grade) and for general skills.

How to use it as a high school or college student:

1. Before a lecture or chapter on a new method, find one problem it solves.
2. Spend 10 to 15 minutes trying, using anything you already know. Write down your attempts.
3. Then read the worked example or attend the lecture and compare the method with your attempts.

Keep the struggle short and always follow it with the real explanation. Unguided struggle with no instruction afterward is just frustration.

The 5 types of study mistakes and how to fix each

Not every mistake needs the same treatment. Rereading the chapter won't fix a careless sign error, and slowing down won't fix a fact you never learned. Label each error first.

If one category dominates your tally, that's your study priority this week.

How to keep an error log

An error log, sometimes called a mistake journal, is a running record of every question you get wrong or guess right, with enough detail to retest it later. It's a staple of LSAT, GMAT, and MCAT prep, but the method works for any course. A spreadsheet, notebook, or doc is fine.

Error log template

What each column is for:

• Source: where the question is, so you can find it again ("Practice exam 2, Q14").
• My answer and confidence: what you put and how sure you were. High-confidence errors and low-confidence correct answers both belong here.
• Correct answer and why: the right answer and the reasoning, in your own words. If you only copy the answer, you'll miss the next variation.
• Type: one of the five categories above.
• Rule for next time: one sentence you'd tell yourself before a similar question.
• Retest dates: two to three days later, then about a week later.

Three filled-in examples

Then turn each entry into a question you can answer from memory: "Where does glycolysis happen?" or "Differentiate sin(5x)." These fit straight into your flashcards or a practice quiz you write yourself.

How to review your error log before an exam

A log only helps if you use it. Work through it on three timescales.

How to learn from mistakes on a test you get back

Do this within two days, while you still remember what you were thinking:

1. Log every lost point, not just the big ones.
2. For each, write what you were thinking at the time. "I thought 'except' meant 'including'" is far more useful than "wrong."
3. Ask the teacher or TA about any question where you still don't understand the correct answer.
4. Schedule retests for every entry over the next week.

Every week: a 20-minute retest

1. Retest, don't reread. Cover the answer columns and redo each open entry from memory.
2. Mark the result. Right from memory: tick. Wrong: add today's date and a new retest date in two or three days.
3. Close entries after two clean retests. Two correct answers from memory on separate days means you can archive it.
4. Tally the types. Count this week's errors by category. If careless slips are half your list, more content review won't help; checking habits will.
5. Write two new questions per pattern. If you keep missing chain-rule problems, write fresh ones rather than redoing the same item until you've memorized it.

The final week before the exam

Redo every open entry once, plus a random sample of closed ones. Spread those retests over several days, not one evening; spacing is what makes them stick, as the spaced repetition guide explains.

It's easy to decide your memory or intelligence is the problem when the tally shows it's mostly misreading or timing. Seeing the pattern in black and white is one of the most practical ways to build a growth mindset, and tracking your own errors is a core skill of metacognition.

Start tonight: open your most recent marked quiz or problem set, log every lost point in the template above, and put the first retest in your calendar for three days from now.