The score at the top of a practice test is the least useful thing on the page. The useful part sits underneath: the questions you got wrong. Each names something you can still fix before the real paper, which no question you answered correctly can do.
Most of that value goes in the next five minutes: a wince at the mark, a glance down the right answers, a closed tab. This guide sets out how to learn from mistakes instead: what to read, how to sort each miss, how to retry it and when to bring it back.
EdProcto's page for students describes a study mode built around the same steps: the reason after every answer, then a list of your misses and a second go at those alone.
Is a wrong answer on a practice test bad for learning?
No. A wrong answer followed by the right one tends to help, and can leave you better off than being handed the answer without trying.
Kornell et al. (2009) arranged for people to fail. Across six experiments they set questions built to be failed, some about invented events and some where a loosely related word had to be guessed. One group tried, then saw the answer. The other saw question and answer together. Those who had tried and failed learned more.
Richland et al. (2009) found the same with an essay on vision: readers questioned on it before reading did better afterwards than readers given extra study time, even counting only the questions they had failed beforehand.
Metcalfe (2017), reviewing the laboratory work, concluded that errors followed by corrective feedback benefit learning, and that it may be worth committing them where the stakes are low. A practice test is that place.
The condition is the correction. Butler and Roediger (2008) gave the answers to a multiple-choice test straight away, later, or never. Against no feedback, both timings raised correct answers at a later recall test and cut the wrong options that resurfaced as answers.
Why do the answers you were sure of matter most?
A correction changes them the most. Butterfield and Metcalfe (2001) predicted the opposite, reasoning that a strongly held wrong answer should get in the way of the right one. At the retest, the errors people had been surest of were the likeliest to be fixed. The effect is called hypercorrection. Hattie and Timperley (2007) pass on one explanation: a learner who expected to be right and is told otherwise studies the item for longer.
The fix does not always last. Butler et al. (2011) asked general-knowledge science questions, took a confidence rating with each answer and showed the correct answers. Hypercorrection appeared in people retested at once and in people retested a week later, but fewer errors were corrected after the week, and where a correction had been forgotten, confident errors were likelier than hesitant ones to reappear.
So flag the misses you would have bet on. EdProcto's multiple choice with confidence asks how sure you are with each answer, so the bet is placed before the result is known.
What should you read after getting a question wrong?
Read the right answer and why it is right. A cross on its own does very little.
Fazio et al. (2010) had 48 undergraduates read passages and answer questions on them. Some answers got no feedback, some got only the word "correct" or "incorrect", and some were followed by the correct answer. Among readers who did not see the passages again, a final test about half an hour later found 18% of the first-time errors put right with no feedback, 31% after "incorrect" alone, and 79% after seeing the answer. The authors' verdict was that right-or-wrong feedback did not help error correction. They were following Pashler et al. (2005), whose 258 participants learned foreign vocabulary: supplying the correct answer after a wrong response increased retention a week later by 494%, while feedback after a correct response made little difference.
So ticks and crosses alone give you little, and correction time belongs on the misses.
The reason counts as well. Butler et al. (2013) gave readers either the correct answer or the answer with an explanation. Two days later both groups did equally well on the repeated questions, but the explanation group did better on new questions that needed inference. An exam is unlikely to repeat your practice question word for word.
How much to read is disputed. Hattie and Timperley (2007) describe an older study in which the plain correct answer beat longer feedback that went through each wrong option. Its authors suspected that dwelling on the errors made them easier to remember. Metcalfe (2017) calls analysis of the reasoning behind the mistake crucial. A fair reading of both: one line on why you chose your answer, then the right one and its reason.
How do you sort the questions you got wrong?
Sort each one into four kinds: a slip, a gap, a misconception or a misread question. The four are a working sort of our own, not a taxonomy from the research, but the line between a gap and a misconception has support. Hattie and Timperley (2007) conclude that feedback is most powerful when it corrects a faulty interpretation, not a total lack of understanding, and that where knowledge is missing, further instruction does more than feedback.
A correction can repair a wrong idea. It cannot stand in for a page you never learned.
| Kind of miss | How to tell | What to do |
|---|---|---|
| Slip | You knew it, and can say at once what went wrong | Name the slip (a sign, a unit, a copying error) and redo the question once |
| Gap | You had nothing to recall, or you guessed | Study the one section the question came from, close it, then answer. A question EdProcto writes from your own material shows the passage behind it |
| Misconception | You were sure, and can give a reason for your wrong answer | Write what you believed, what is true and why. Then answer a new question on the same idea |
| Misread question | Your answer is right for a different question | Find the word you skipped (NOT, two, least, the units) and redo it. Nothing needs relearning |
A right answer that was a guess goes on the list too, as a gap.
A worked example: four misses on one physics quiz
A twelve-question quiz on forces and motion comes back 8 out of 12.
- Question 3. "A 2 kg trolley accelerates at 3 m/s². What is the resultant force?" Answered 5 N. The rule was known, but the numbers were added. A slip. Force is mass times acceleration: 6 N.
- Question 5. "What is the unit of momentum?" Guessed "joule". A gap. The answer is the kilogram metre per second.
- Question 8. "A ball is thrown straight up. Ignoring air resistance, what is the resultant force on it at the top of its flight?" Answered "zero", with certainty, believing that an object at rest has no force on it. A misconception. The force is the ball's weight, acting downwards. Its velocity is zero for an instant, yet it is still accelerating downwards at about 9.8 m/s².
- Question 11. "Which of these is NOT a vector: velocity, force, speed, displacement?" Answered "force", having read it as "which is a vector". A misread question. The answer is speed.
Four misses, four different jobs. Question 8 deserves ten minutes. Question 11 deserves ten seconds.
Why answer again from memory instead of reading the correction?
Because following a correction on the page is easier than producing it, and the exam will ask you to produce it. In Kornell et al. (2009), seeing the answer without first attempting it was the weaker condition.
Rawson and Dunlosky (2011) built a method on the second attempt. Their students tried to recall each concept, restudied it, and practised again until it was recalled correctly, more than 100,000 written answers in all.
For one sitting:
- Read every correction first, then cover them.
- Answer each miss in writing, in full. "Oh yes, of course" does not count.
- For a gap or a misconception, add the reason the right answer is right.
- Check. If you miss again, reread the correction, do two or three other questions, then try once more.
- Stop when every miss has been answered correctly from memory, ideally three times. In EdProcto's study mode this step ends each attempt: your misses are listed, and the retry holds only those questions.
A mastery loop repeats a missed idea with a similar question until it is right.
When should each wrong answer come back?
Bring each miss back the same day, then on three later days spaced well apart. Same-day success proves little: the correction is minutes old.
Rawson and Dunlosky (2011) tested their students one to four months on. The number of correct recalls on the first day mattered less and less as later sessions were added, while those later sessions had a pronounced effect on retention for little extra practice. Their recommendation: recall a concept correctly three times on the first day, then relearn it on three later occasions set well apart.
The studies fix no schedule to the day. One of our own: the misses again two or three days later, the whole quiz about a week after that, and once more before the exam.
EdProcto keeps the diary: the quiz falls due again after a few days, and your dashboard shows what is due and what you missed last time. Two neighbouring guides cover the first sitting on a chapter and when something counts as known.
How do you keep an error log that takes two minutes?
Write one line for each miss: the question in a few words, the kind of miss, the right idea in your own words, and the date it is next due. No study we read tested an error log as such. It makes sure the three things that were tested do happen: the correction, the second attempt and the return visit.
| Question | Kind | The right idea | Next due |
|---|---|---|---|
| Q3, force on a trolley | Slip: added | F = m × a, so 2 × 3 = 6 N | 12 Oct |
| Q5, unit of momentum | Gap | Mass × velocity, so kg m/s | 12 Oct |
| Q8, ball at the top | Misconception | Still its weight, downwards. Not moving is not the same as no force | 12 Oct |
Write the right idea, not the wrong answer at length. Tick a line each time you get it right on a later day, and retire it after three ticks. Read down the "kind" column once a week: mostly slips means slow down and check, and mostly gaps means you tested before you had studied.
What should you not do with wrong answers?
Do not stop at the score, reread the whole chapter, or redo the questions you already got right.
- Counting the score and stopping. The score says how many, and the misses say which.
- Rereading the whole chapter. Going back to the text does correct errors: in Fazio et al. (2010), readers who reviewed the passages put right 68% of the errors that had received no feedback. But readers shown the correct answer, with no review, put right 79%, and the reviewers spent about as long on sentences they had answered correctly (3.2 seconds) as on sentences they had got wrong (3.4 seconds). With a vocabulary list, in a second experiment, people did linger on the words they had missed. Reread the section behind each gap, and only that.
- Redoing what you got right. It feels good and teaches little that day. Those questions return when the whole quiz falls due.
How does EdProcto handle the questions you got wrong?
EdProcto is the best tool for this job when the practice test comes from your own material, because its study mode is built on the three things the research above supports: the correct answer with its reason, a second attempt from memory, and a return visit.
Paste a chapter or your notes into the quiz maker on the home page and the questions are written from them by Procto AI. In study mode every question is followed by its answer and a plain explanation, so no miss is left as a bare cross. Each attempt closes with what you missed and a retry of those questions alone. Days later the quiz is due again. A whole book can be taken chapter by chapter, with a map of the chapters you hold.
With nothing to paste, take a ready-made quiz from the kits.
The Free plan includes study mode, the re-tests, the kits and 20 questions a month from Procto AI. The Learner plan lifts the limit on self-tests, raises the written questions to 100 a month and adds the whole book with its map. See the pricing page. Deciding what kind of miss you made stays yours, and no tool can promise a grade.
Quick questions
Should I retake the whole practice test or only the questions I got wrong?
Only the misses on the same day, then the whole test when it falls due. EdProcto's page for students explains how the retry and the re-test work.
Is it bad to guess on a practice test?
No. Kornell et al. (2009) found that a failed attempt followed by the answer taught more than reading the two together. Log a lucky guess as a gap.
How many times should I get a missed question right before I drop it?
On more than one day. Rawson and Dunlosky (2011) recommend three correct recalls on the first day, then three later sessions spaced well apart.
What if I keep getting the same question wrong?
Write out your reasoning step by step and find the step that fails. A miss that survives two corrections may hide a gap further back, and is worth showing to a teacher.
References
- Butler, A. C., Fazio, L. K., & Marsh, E. J. (2011). The hypercorrection effect persists over a week, but high-confidence errors return. Psychonomic Bulletin & Review, 18(6), 1238–1244. https://doi.org/10.3758/s13423-011-0173-y
- Butler, A. C., Godbole, N., & Marsh, E. J. (2013). Explanation feedback is better than correct answer feedback for promoting transfer of learning. Journal of Educational Psychology, 105(2), 290–298. https://doi.org/10.1037/a0031026
- Butler, A. C., & Roediger, H. L. (2008). Feedback enhances the positive effects and reduces the negative effects of multiple-choice testing. Memory & Cognition, 36(3), 604–616. https://doi.org/10.3758/MC.36.3.604
- Butterfield, B., & Metcalfe, J. (2001). Errors committed with high confidence are hypercorrected. Journal of Experimental Psychology: Learning, Memory, and Cognition, 27(6), 1491–1494. https://doi.org/10.1037/0278-7393.27.6.1491
- Fazio, L. K., Huelser, B. J., Johnson, A., & Marsh, E. J. (2010). Receiving right/wrong feedback: Consequences for learning. Memory, 18(3), 335–350. https://doi.org/10.1080/09658211003652491
- Hattie, J., & Timperley, H. (2007). The power of feedback. Review of Educational Research, 77(1), 81–112. https://doi.org/10.3102/003465430298487
- Kornell, N., Hays, M. J., & Bjork, R. A. (2009). Unsuccessful retrieval attempts enhance subsequent learning. Journal of Experimental Psychology: Learning, Memory, and Cognition, 35(4), 989–998. https://doi.org/10.1037/a0015729
- Metcalfe, J. (2017). Learning from errors. Annual Review of Psychology, 68, 465–489. https://doi.org/10.1146/annurev-psych-010416-044022
- Pashler, H., Cepeda, N. J., Wixted, J. T., & Rohrer, D. (2005). When does feedback facilitate learning of words? Journal of Experimental Psychology: Learning, Memory, and Cognition, 31(1), 3–8. https://doi.org/10.1037/0278-7393.31.1.3
- Rawson, K. A., & Dunlosky, J. (2011). Optimizing schedules of retrieval practice for durable and efficient learning: How much is enough? Journal of Experimental Psychology: General, 140(3), 283–302. https://doi.org/10.1037/a0023956
- Richland, L. E., Kornell, N., & Kao, L. S. (2009). The pretesting effect: Do unsuccessful retrieval attempts enhance learning? Journal of Experimental Psychology: Applied, 15(3), 243–257. https://doi.org/10.1037/a0016496