The quiz is marked and a third of it went badly. A reteaching plan from one quiz is half a page that says which misses get lesson time, for whom, taught how, and how you will know it worked.
Most of the work is sorting each weak result onto one of three lists: reteach to everyone, reteach to a group, or leave it. Some belong on none, because the question failed and the class did not.
EdProcto, which publishes this guide, prints cautions about a quiz's figures on every plan's results and, in its full analysis, names the topics to reteach. What to reteach tomorrow covers reading that page; this guide runs from there to the re-check. Sources were read on 10 October 2026.
What goes in a reteaching plan made from one quiz?
Five things: what everyone meets again, who sits in a group, what is left alone, what changes in the lesson, and the date of the second quiz.
- Clear the figures. Grade what is waiting and read the cautions on EdProcto's results.
- Rule out the quiz. Check the question, the clock and the number of sittings before a miss counts as a gap.
- Sort. Each weak topic or question goes to everyone, to a group, or is left.
- Write the change. One line a reteach: how it was taught, how it will be taught now.
- Book class time. Minutes in a named lesson, not homework.
- Write the second quiz now. New questions on the retaught ideas, for a few days after the reteach.
What did the research find about acting on a quiz, and how often do teachers act?
In the studies we read on responding to results, from formative assessment, mastery learning and data-driven reteaching, most teachers said they act, the response was often the same content taught the same way, and achievement moved when the teaching changed too. We read abstracts unless we say otherwise.
How often teachers act, and what they do
- Most say they adjust. Hoover and Abrams (2013) surveyed teachers in one large suburban Virginia district about their summative assessment data. A majority reported using results to judge and adjust their teaching. Quick analysis was fairly regular, and closer analysis less so.
- A classroom quiz changes more. Farrell and Marsh (2016) spent a year in five middle schools. District benchmark results went with reteaching, retesting and small groups, usually with no change of method. Classroom assessment results and student work were, in proportion, most often tied to a change in how teachers taught.
We found no count of how often one quiz leads to one changed lesson.
What moved achievement
Fuchs et al. (1991) assigned 33 teachers at random, for 20 weeks of mathematics, to no measurement, regular measurement of progress, or measurement with advice from an expert system. Each followed two pupils with learning or emotional difficulties. Both measuring groups appeared to revise their teaching more often than the control group, but only the advised group's pupils achieved more. Its teachers changed which problems they taught and how; the others changed only which problems to reteach.
Correctives in mastery learning
Mastery learning follows each short test with corrective instruction for the students below a standard. What to reteach tomorrow has its pooled studies and the dispute over them.
- Different, and in class. Guskey (2007), read in full, reports Bloom's condition: correctives work only if they are "qualitatively different from the original instruction". Set as homework they rarely succeed, because the students who need them most are the least likely to do them.
- Someone pays in time. Arlin and Webster (1983) assigned 88 seventh-grade students at random to a mastery or a non-mastery approach over four chapters. The mastery group achieved more, and the price seemed to be extra remedial time and time the faster learners lost.
No study here tested one teacher's plan from one quiz, so the rules below are ours.
How do you sort results into everyone, a group and leave it?
Sort by who missed it: most of the class, the same few students, or a scatter of different ones.
| What the results show | List |
|---|---|
| Most of the class missed it, and one wrong answer leads | Reteach to everyone |
| The same few students missed two or more questions on a topic the rest hold | Reteach to a group, with worthwhile work for the rest |
| Different students missed different questions, a few each | Leave it: a word at the desk |
| A topic the class holds | Leave it, and drop the recap |
| The question failed, the time ran out, or answers wait for a grade | No list: put the quiz right first |
"Everyone" is not the safe choice, since those who had it pay in time (Arlin and Webster, 1983). Break ties by what comes next, a rule of ours: a middling result goes to everyone if the next unit stands on it, and can be left if the topic will not return.
Count students as well as percentages
A topic's percentage hides who. In EdProcto a topic's score is the share of the topic's points the class earned, so a question worth three points counts three times and can pull a topic down alone. On every plan, "Download the marks" gives a row for each attempt and a column for each question.
How do you tell a gap in the teaching from a faulty question before you reteach?
Rule out four things that look like a gap and are not: too few sittings, the clock, answers not yet graded, and a question that failed. EdProcto prints a caution for the first three on every plan.
| It looks like | It may be | How to tell |
|---|---|---|
| A weak topic in a group of four | Too few sittings | "Fewer than five submissions: the numbers below describe these students, not the class or the teaching yet." |
| Low scores on the last questions | The clock | A caution on EdProcto's results names the questions some students never reached. Each is zero in their scores, and its "Got it" is of the students who did reach it |
| An open question near zero | Answers not yet graded | "3 answers are waiting to be graded. Until then they count as zero in the scores above, so grade before you read them." |
The full analysis prints "Reteach before moving on" for a topic under 60% and "Secure, and safe to build on" at 85% or above. Both are said only from five sittings, and a topic most of the class never reached is named in neither.
Working a suspect question
A question most of the class missed may have a wrong key or two defensible answers. Answer it yourself without the key, then ask who chose what. EdProcto's full analysis prints "Check the key" where the stronger students did worse on a question than the weaker ones, or where a wrong option drew more students than the key and they did better on the rest of the quiz. A popular wrong answer without those signs is filed under "A shared wrong answer". The question with the wrong key has the full check, and finding misconceptions tells a shared wrong idea from a slip.
Two more tests. Ask three students who missed it to talk through their answer: if they have the idea and read the question another way, the question failed. And check the register: students absent for the lesson are a group.
What should a reteach change?
The way the idea is presented, what students do with it, and where it happens. Which content to go over is the easy half: in Fuchs et al. (1991), the pupils of teachers who changed only that did no better than the control group's.
| First time | In the reteach |
|---|---|
| The rule stated, then examples | Examples first, the rule drawn from them |
| Symbols on the board | A diagram, a table or objects to count |
| The teacher explains | Pairs explain, starting from the popular wrong answer |
| Practice at home | Practice in the room, checked as it is done |
Guskey (2007) gives three marks of a good corrective: it presents the idea differently, it has students do something different, and it lets them succeed. If it is not working, drop it for another. (He grounds the second in learning styles, which we leave aside.)
Change at least one row and say which in the plan. Students who had it the first time need something worth doing: in Guskey's example they wrote word problems in pairs or chose to tutor a classmate.
How do you check the reteach worked with a short second quiz?
Set a short second quiz of new questions on the retaught ideas a few days after the reteach, and decide before it is sat what result will count.
- New questions, same ideas. Winget and Persky (2022), a review for pharmacy educators whose full text we consulted, describe a second, parallel assessment: the same objectives and standard, in questions that may differ in form. The first quiz sat again measures memory of it.
- Three to five questions for each idea. Our working rule; see how many questions a quiz needs.
- A few days later. A right answer ten minutes after the explanation shows it was heard. Two to four days is our working rule.
- A new quiz, not the old one reopened. EdProcto reads a class on first sittings: a later sitting of the same quiz under the same email address is listed under "Students" and left out of the figures. Build the re-check as its own quiz, in multiple choice or another shape; Procto AI can draft the questions from your notes for you to check.
- The standard, written first. For example, three of four right for each student. The review puts the usual mastery standard between 80% and 90%.
- Read it by name. Students still short get ten minutes with you. If most of the class is still short, the reteach failed: change the approach again.
A second quiz cannot prove the reteach caused a rise: the new questions may be easier, and on four questions one answer is 25 points. Pre-test and post-test says what two scores can show. The review also recommends cumulative assessments, so ask the idea again in the next unit's quiz.
A worked example: one probability quiz and the plan it gave
This class is an illustration: its students and answers are invented, and the figures and quoted sentences are what EdProcto's analysis code returned for them. Twenty-six students had twenty minutes for eleven multiple-choice questions on probability, one a page, Q9 worth three points and the rest one.
| Topic | The page printed | After the checks | List |
|---|---|---|---|
| Combined events, Q7 to Q9 | 50.8% | The key on Q9 stands | Everyone |
| Listing outcomes, Q4 to Q6 | 59% | Q6 failed. Without it, 40 of 52 right (76.9%), and five students missed both Q4 and Q5 | A group of five |
| Expected frequency, Q10 and Q11 | 80.5% | Eight students never reached Q11 | Leave it |
| Single events, Q1 to Q3 | 91% | Nothing | Leave it |
The page said "Reteach before moving on: Combined events (50.8%), Listing outcomes (59%)." The checks moved one of them.
- Q9, a real gap. A spinner lands on red with probability 0.3: what is the probability of two reds in two spins? Nine students chose 0.09 and averaged 91.4% on the quiz. Fourteen chose 0.6, adding where they should multiply, and averaged 57.1%: "A shared wrong answer on Q9". "Got it" on Q7 to Q9 read 81%, 69% and 35%, which average 62%, but Q9 carries three of the topic's five points, hence 50.8%.
- Q6, a failed question. Two dice are thrown and the scores added: how many outcomes are there? The key was 36. Fifteen students, the highest scorers among them, chose 11, the number of possible totals: "Check the key on Q6: the stronger students got it wrong more often than the weaker ones." Both answers can be defended.
- The clock. The caution named Q11 (8) and Q10 (3). "Got it" on Q11 was 78% of the eighteen who reached it; with the eight counted as wrong it would read 54%.
The first lesson had stated the rule "and means multiply" with two examples. The reteach, twenty minutes for everyone, used a ten-by-ten grid: three columns shaded for the first red, three rows for the second, nine squares shaded twice. The class then tested 0.6 against sense: two reds cannot be likelier than one. The five then listed the outcomes of a coin and a die with the teacher while the rest practised.
Three days later twenty-five students sat seven new questions. Combined events read 78%, the adding option took 22 of 100 answers, and the page said "Secure, and safe to build on: Listing outcomes." Five students had two or fewer of four right on combined events: ten minutes each, not a third lesson. The rise is a reason to move on, not a measurement of the reteach.
How does EdProcto turn one quiz's results into the plan?
EdProcto is the best tool for this job when the plan has to be right before it is taught, because its results page puts the cautions first and, in the full analysis, says in a sentence what to reteach. In the example one page showed whom the clock cut short, pointed at Q6 and named two topics; a bare list of percentages would have sent both topics to the whole class.
- On every plan. The average and its spread, the median and the pass rate; each student's score; "Each question" with "Got it", the share of its points the class earned; the cautions, which below Educator stand under "About these figures", "Read this first"; and "Download the marks".
- In the full analysis, on Educator and School. "By topic", weakest first; the reteach and secure sentences; "A shared wrong answer" and "Check the key"; each option's "Chose it" and "Their mean".
Free takes thirty students a quiz and ten quizzes a month, the re-check among them. Tutor has no limit on quizzes and takes 100 students; Educator takes 500 and adds the full analysis. See pricing and the pages for teachers and for universities. A score is not a grade.
Quick questions
How is corrective instruction different from reteaching?
Corrective instruction is mastery learning's term for teaching a missed idea again, differently, to the students who missed it. Reteaching is one corrective among several in Guskey (2007).
How many students must miss a question before I reteach it to everyone?
No study we read sets a number. Our working rule is most of the class, with one wrong answer leading.
Should the second quiz repeat the first quiz's questions?
No. Use new questions on the same ideas, so a right answer shows the idea, not a memory of the first quiz.
Does EdProcto write the reteaching plan for me?
No. On the Educator plan a quiz's results name each topic under 60% once five students have sat it, and every plan gets each question's share right and the cautions. The lists, the lesson and the second quiz are yours.
References
- Arlin, M., & Webster, J. (1983). Time costs of mastery learning. Journal of Educational Psychology, 75(2), 187–195. https://doi.org/10.1037/0022-0663.75.2.187
- Farrell, C. C., & Marsh, J. A. (2016). Metrics matter: How properties and perceptions of data shape teachers' instructional responses. Educational Administration Quarterly, 52(3), 423–462. https://doi.org/10.1177/0013161X16638429
- Fuchs, L. S., Fuchs, D., Hamlett, C. L., & Stecker, P. M. (1991). Effects of curriculum-based measurement and consultation on teacher planning and student achievement in mathematics operations. American Educational Research Journal, 28(3), 617–641. https://doi.org/10.3102/00028312028003617
- Guskey, T. R. (2007). The rest of the story. Educational Leadership, 65(4). https://www.ascd.org/el/articles/the-rest-of-the-story
- Hoover, N. R., & Abrams, L. M. (2013). Teachers' instructional use of summative student assessment data. Applied Measurement in Education, 26(3), 219–231. https://doi.org/10.1080/08957347.2013.793187
- Winget, M., & Persky, A. M. (2022). A practical review of mastery learning. American Journal of Pharmaceutical Education, 86(10), Article ajpe8906. https://pmc.ncbi.nlm.nih.gov/articles/PMC10159400/