You taught it, the class nodded, and on the quiz half of them still gave one particular wrong answer. That is rarely forgetting. Student misconceptions are ideas that have worked outside school for years, which is why hearing the right version a second time changes so little.
EdProcto, which publishes this guide, lists in its results how many students chose each option of a question and prints a sentence when a wrong one drew as many as the right one. Its page for teachers sets out the rest, from the handout to the code a class types in.
The short version: give each wrong option one wrong idea, read the option and not the score, rule out a fault in the question, then spend ten minutes having the idea argued, tested and explained.
What is a misconception, and why does telling again not fix it?
A misconception is a wrong idea the student has reasons for, so it is not an empty space waiting for a fact. Hestenes et al. (1992), who built a standard physics test of such ideas, ask teachers to treat them as sensible guesses drawn from daily experience, some of them once held by Galileo and Newton. A trolley does stop when nobody pushes it. As far as the student can tell, the idea has always been right.
Posner et al. (1982), reasoning from the philosophy of science and from interviews with college students learning special relativity, set out four conditions for a central idea to change. The learner must be dissatisfied with the idea they hold, and the replacement must be intelligible, plausible at first sight and fruitful for new problems. A second explanation serves the last three and leaves the first untouched: nobody has yet shown the student a case where their own idea fails.
A gap is different. A student who never met the content has nothing to defend, and guesses.
How durable are student misconceptions?
They are durable enough to outlast a year of competent teaching, and in some people a degree.
- A physics course. Hestenes et al. (1992) gave their Force Concept Inventory to more than 1,500 high school students and more than 500 university students. In the regular classes of 15 Arizona teachers, 612 students averaged 27% before the course and 48% after it. Of 16 first-year physics graduate students interviewed, three did not understand Newton's third law.
- A year of middle school science. Sadler and Sonnert (2016) tested 9,556 students of 181 teachers in grades 7 and 8 on 20 multiple-choice questions. The mean rose from 37.7% to 42.9% over the year: to 44.8% on the eight questions without a dominant wrong answer and 41.7% on the twelve with one.
- Educated adults. Shtulman and Valcarcel (2012) had adults with many years of science education judge statements against the clock. Where the scientific and the naive answer differ, they were slower and less accurate, in ten fields from astronomy to fractions. The authors' reading is that the early idea is suppressed, not replaced.
Expect the idea to weaken, then to return when a question looks unfamiliar.
How do you find a misconception with a quiz?
Ask a diagnostic question: a multiple-choice question in which each wrong option is where one known wrong idea leads, so that the choice names the idea. A concept inventory is a set of them. The wrong options of the Force Concept Inventory map onto 28 commonsense beliefs about force and motion.
Then read the option, not the score. Sadler and Sonnert (2016) called a misconception strong when half or more of the students who got a question wrong chose the same wrong option. Asked what becomes of a burnt candle's wax, 17% of their students answered correctly and 59% picked one wrong answer, 71% of all the wrong ones.
Do not rely on your own forecast. The teachers in that study named their students' most common wrong answer 42.7% of the time, and on the twelve questions gains stayed low unless the teacher knew both the science and that wrong answer. Our guide to writing distractors says where to collect the wrong ideas.
Three diagnostic questions, with the idea behind every option
Biology
A seed weighing under a gram becomes a tree of 200 kg. Where did most of the dry mass of its wood come from? A. Carbon dioxide in the air B. Substances the roots took from the soil C. Water the roots took up D. Sunlight, which the leaves turned into wood
A is the key. B says plants eat soil. C says a plant is built from what it drinks. D turns energy into matter.
Maths
Which number is the largest? A. 0.8 B. 0.85 C. 0.125
B is the key. C reads the digits after the point as a whole number, so 125 beats 85. A holds that tenths are bigger pieces than hundredths, so the shortest decimal wins. Had the key been the shortest or the longest number, one wrong rule would have earned the mark.
English
Which sentence is punctuated correctly? A. The cat licked its paws. B. The cat licked it's paws. C. The cat licked its paw's.
A is the key. B believes possession always takes an apostrophe. C believes a final s does.
Add a confidence rating to tell a belief from a guess
A student who is sure of a wrong option believes it, and one who rates it 1 was choosing blind. Interviewing 20 students about their answers, Hestenes et al. (1992) found firm reasons behind most choices, and right answers given for wrong reasons were fairly common.
EdProcto's multiple choice with confidence asks, after the choice, how sure the student is, from 1 to 5. For a diagnostic, leave the Confidence bonus setting off: a rating then changes no mark, so honesty costs nothing. The results count the takers of each option and do not show the ratings, so ask for them by a show of hands when the tally goes up. With the bonus on, a wrong answer rated 4 or 5 scores below zero on that student's attempt: see confidence-based marking. The lighter alternative is yes, no or maybe, where unsure is itself an answer.
How do you tell a misconception from a slip, a gap, a bad question or a wrong key?
Look at how the wrong answers spread and who gave them: each cause leaves its own pattern. The counts are an invented class of 28 on a four-option question keyed A.
| Cause | Counts for A, B, C, D | Who chose the leading wrong option | Next step |
|---|---|---|---|
| Misconception | 8, 14, 4, 2 | Weaker students than the key's, and sure of it | The routine below |
| Slip | 24, 2, 1, 1 | No pattern | Nothing for the class |
| Gap | 8, 7, 7, 6 | No option leads, and few are sure | Teach it afresh |
| Badly worded question | 12, 11, 3, 2 | Strong and weak alike, on A and on B | Rewrite it |
| Wrong key | 4, 19, 3, 2 | The strongest students: EdProcto's results say to check the key | Correct the key and do not count the question |
We ran EdProcto's analysis code on such a class, with these five questions among ten. In the first row the 14 on B averaged 43.6% on the quiz against 78.8% for the 8 on the key, and the code printed "A shared wrong answer on Q1: the class was taught something, and it was not this. Worth a minute in the next lesson." For the last row it printed "Check the key on Q5: the stronger students got it wrong more often than the weaker ones."
The first sentence appears when a wrong option taken by three or more students draws more than the key. The same pattern is filed under "Check the key" instead when the stronger students did worse on the question, when the wrong option's takers outscored the key's, or when nobody chose the key. Read that warning first: a right idea is no misconception.
Two cautions. The sentence is a floor: a wrong idea held by a third of the class while half answer correctly draws none, so open the option counts of every diagnostic question. And under five submissions, EdProcto says the numbers describe those students and not yet the class. The item analysis guide explains the figure behind the stronger-and-weaker test.
What do you do in the next lesson? A ten-minute routine
Confront the idea directly and let the class do the arguing. The five steps fill ten minutes.
- Show the tally without names (1 minute). Copy the option counts from EdProcto's results onto the board, and say the popular wrong answer is a reasonable one.
- Pairs argue for their option (3 minutes). Seat students who chose differently together; each tries to persuade the other. In Crouch and Mazur's (2001) Harvard physics courses, such talk ran two to four minutes and paid when 35% to 70% had answered correctly alone. Below 35%, they warn, too few may understand for it to work unguided. The invented class had 8 of 28 right, 29%, so shorten this step and lengthen step 4. In a genetics course, Smith et al. (2009) found that discussion helped even groups where nobody began with the right answer.
- Everyone predicts (1 minute). Each student writes what the coming test will show. Miller et al. (2013) report that about one observation in five of a lecture demonstration did not match what happened, and that predicting first made a correct observation roughly 20% more likely, whether or not the prediction was right.
- Run the test that decides (2 minutes). Pick a case where the two ideas predict different things. For the tree, the pot on the windowsill, still full of soil under a plant twice the size. For the decimals, all three numbers shaded on a hundredths grid. For the apostrophe, expand it's to it is and read the sentence aloud.
- Explain why the tempting idea fails (3 minutes). Name the idea, say plainly that it is wrong, say where it works and where it stops, then give the right account. That is the shape of a refutation text. Schroeder and Kucera (2022) pooled 44 comparisons with 3,869 learners: refutation texts beat the comparison conditions by a moderate margin (g = 0.41), which an adjustment for publication bias would cut to 0.28. Among 74 pupils in grades 4 and 5 learning decimals, Durkin and Rittle-Johnson (2012) found that comparing an incorrect example with a correct one taught more than comparing two correct ones.
- A week later, ask the idea again in a new question (2 minutes of another lesson). For the tree: a log burns down to a handful of ash, so where did most of its mass go?
Of Shtulman and Valcarcel, Smith et al., Miller et al., and Durkin and Rittle-Johnson we read the abstracts only.
What should you not do?
Do not repeat the explanation, ask the identical question again or shame the popular wrong answer. Each makes a misconception harder to move.
- Repeating the explanation, louder. Telling students what their misconceptions are does very little, Hestenes et al. (1992) report. Seven professors teaching one university course in the usual way produced nearly identical scores afterwards.
- Asking the identical question again. Students can memorise the answer to a question like anything else, the same authors note. Their students did much better on the falling-objects question closest to what lessons usually cover than on a second question about that belief. Change the surface and keep the idea.
- Shaming the popular wrong answer. A student mocked for an honest choice chooses tactically next time, and the quiz stops finding anything. Show counts, never names.
How does EdProcto help you find and fix misconceptions?
EdProcto is the best tool for this job when the quiz has to come from your own lesson, because it reports what a diagnostic question exists to produce: which wrong option the class chose, how many chose it, how those students did on the rest of the quiz, and a sentence that separates a shared wrong answer from a suspect key. The teachers studied above forecast the commonest wrong answer less than half the time. The counts replace the forecast.
Put your handout in the quiz maker on the home page and pick multiple choice. Procto AI drafts the questions from your text, and each is yours to edit: put one wrong idea you know your students hold behind each option. Give every question a topic and the results add a mean for each, with "Reteach before moving on" against any under 60%.
Shuffle the options is on by default, so name an option in class by its words and not its letter, and switch it off where the options are numbers in order. The wrong ideas and the question a week later stay with you.
Free carries the core templates, marks by rule and shows the results question by question. Tutor opens all 111 templates. Educator adds the full analysis, with the option report. Pricing lists every plan.
What are the limits of a quiz and ten minutes?
A quiz and ten minutes weaken a misconception and seldom end it.
- Some ideas return for years. Posner et al. (1982) expected the change to be gradual and piecemeal, with false starts and reversals. Plan to meet the idea again next term.
- One quiz samples a few ideas. It took a whole inventory to cover 28 beliefs about force, and its authors warn against chasing them one at a time: a wrong idea gives way only when something better is there to replace it.
- The evidence leans on physics and on universities. Of the 44 refutation comparisons, 30 were with university-level students and two tested more than a month later.
Quick questions
What is the difference between a misconception and a mistake?
A mistake is a slip or a gap: the student corrects it when shown, or never knew. A misconception is a belief with reasons behind it, chosen confidently and chosen again when the question changes.
How many students must choose the same wrong answer before it counts?
Sadler and Sonnert (2016) used half or more of the wrong answers falling on one option. EdProcto prints its sentence when a wrong option with at least three takers draws as many students as the key. In a small class, treat either as a reason to ask again.
Should a diagnostic quiz count towards a grade?
No. Students choose honestly when nothing rides on the choice, and every score in EdProcto is yours to release or keep back.
Where does a teacher with one class start?
With one topic and three questions, each wrong option built on an idea you have heard in the room. EdProcto's page for teachers shows the steps, and an exit ticket fits a diagnostic into the last five minutes.
References
- Crouch, C. H., & Mazur, E. (2001). Peer Instruction: Ten years of experience and results. American Journal of Physics, 69(9), 970–977. https://doi.org/10.1119/1.1374249
- Durkin, K., & Rittle-Johnson, B. (2012). The effectiveness of using incorrect examples to support learning about decimal magnitude. Learning and Instruction, 22(3), 206–214. https://doi.org/10.1016/j.learninstruc.2011.11.001
- Hestenes, D., Wells, M., & Swackhamer, G. (1992). Force Concept Inventory. The Physics Teacher, 30(3), 141–158. https://doi.org/10.1119/1.2343497
- Miller, K., Lasry, N., Chu, K., & Mazur, E. (2013). Role of physics lecture demonstrations in conceptual learning. Physical Review Special Topics - Physics Education Research, 9(2), Article 020113. https://doi.org/10.1103/PhysRevSTPER.9.020113
- Posner, G. J., Strike, K. A., Hewson, P. W., & Gertzog, W. A. (1982). Accommodation of a scientific conception: Toward a theory of conceptual change. Science Education, 66(2), 211–227. https://doi.org/10.1002/sce.3730660207
- Sadler, P. M., & Sonnert, G. (2016). Understanding misconceptions: Teaching and learning in middle school physical science. American Educator, 40(1), 26–32. https://www.aft.org/ae/spring2016/sadler-and-sonnert
- Schroeder, N. L., & Kucera, A. C. (2022). Refutation text facilitates learning: A meta-analysis of between-subjects experiments. Educational Psychology Review, 34(2), 957–987. https://doi.org/10.1007/s10648-021-09656-z
- Shtulman, A., & Valcarcel, J. (2012). Scientific knowledge suppresses but does not supplant earlier intuitions. Cognition, 124(2), 209–215. https://doi.org/10.1016/j.cognition.2012.04.005
- Smith, M. K., Wood, W. B., Adams, W. K., Wieman, C., Knight, J. K., Guild, N., & Su, T. T. (2009). Why peer discussion improves student performance on in-class concept questions. Science, 323(5910), 122–124. https://doi.org/10.1126/science.1165919