# Homework for tutoring students: a quiz that marks itself

**How should a tutor set homework between lessons?** Homework for tutoring students works best as one short online quiz that marks itself, done alone two to four days after the lesson: four to ten questions by age, most on last lesson's point, one older point and one step ahead. The tutor reads the misses before the next lesson and plans its first ten minutes from them.

Published 2026-10-10 by EdProcto. https://edprocto.com/guides/between-lessons-homework-quizzes-that-mark-themselves

Homework for tutoring students has a job that school homework does not: it must tell you, before the next lesson, what the last one left behind. A worksheet handed over at the door tells you at the cost of the first quarter of an hour.

A short online quiz, done alone in mid-week and marked by rule, tells you the evening before. EdProcto, which publishes this guide, builds its page [for tutors](https://edprocto.com/solutions/tutors) on that loop. The check inside the lesson is a different job, covered in [the tutor's weekly check](https://edprocto.com/guides/the-tutors-weekly-check-a-four-minute-quiz).

## Does homework improve results, and where do the studies disagree?

On balance it does, modestly, but the evidence disagrees about age. Cooper et al. (2006) reviewed United States research from 1987 to 2003, found design flaws in every study, and still found generally consistent evidence that homework helps achievement. The link was stronger in grades 7 to 12 than from kindergarten to grade 6.

Fan et al. (2017), pooling thirty years of maths and science studies, found a small positive relationship, stronger for elementary and high school students than for middle school students. That does not fit a rule that homework pays more with each year of age. Both reviews describe school homework, not a tutor with one student.

### Time spent against homework done

Minutes are the wrong measure. In three studies by Trautwein (2007), homework being assigned went with achievement at the level of the class, and doing it went with gains at the level of the student, but time on homework captured neither.

Fernández-Alonso et al. (2015) analysed 7,451 Spanish adolescents. Prior knowledge outweighed every other variable combined. Of the homework variables, how often it was set and how independently the student worked mattered most, ahead of effort and time, and the best duration was an hour a day. So set it weekly, make it doable alone, and judge it by what was done.

### Online homework with feedback as the student works

The format by itself does little. Of the 31 studies reviewed by Magalhães et al. (2020), most with college students, half found no difference between online and traditional homework, nine favoured online, one favoured traditional and six were mixed.

One positive result came with more than a format. In the randomised trial of Roschelle et al. (2016), with 2,850 seventh-grade students in 43 Maine schools, an online tool gave students feedback and hints as they did maths homework and gave teachers organised information about the work, and the teachers were trained. End-of-year test scores were significantly higher than in a control group that kept its existing homework, and students with low prior achievement gained most. Our inference: the benefit comes when somebody reads the result and teaches differently.

## Why is a quiz between lessons retrieval practice and spacing at once?

Because it makes the student recall the lesson, and does so after a gap: two separate effects from one piece of work. In Roediger and Karpicke (2006), students who re-read a passage did better five minutes later, while students who had been tested on it remembered substantially more two days and a week later.

Rawson and Dunlosky (2022) review the evidence for doing both, which they call successive relearning: practise until the answer is right, then again in later sessions spaced apart. A lesson, a quiz three days on and a check at the next lesson is that schedule built for the student. Neither paper studied tutoring, so the size of the gain for one student is unknown.

## How much homework should a tutor give at each age?

One quiz a week, of four to ten questions by age, is our working rule: no study we found sets a figure for a tutor's homework. It sits on top of the school's, so keep it small.

| Age | Questions | Minutes, roughly |
| --- | --- | --- |
| 7 to 10 | 4 | 3 |
| 11 to 13 | 6 | 6 |
| 14 to 16 | 8 | 10 |
| Over 16 | 10 | 15 |

The research supports the shape of the table, not its numbers: the link with achievement was weaker for younger children in Cooper et al. (2006), and frequency beat duration in both studies of time above.

Ask for the quiz two to four days after the lesson (our rule too). The same evening it tests an echo, and the hour before the next lesson leaves no time to read it.

## What should a homework quiz ask?

Ask mostly about last lesson's point, then one older point and one step ahead. In six questions that is four, one and one.

- **Last lesson's point, four ways.** Not four copies: the idea used forwards, backwards, in words and with awkward numbers.
- **One older point.** Something from three or four weeks back that the student got right then.
- **One step ahead.** A question on what you will teach next, so you know where that lesson can start. This slot is our own choice. Tell the student it has not been taught yet.

Prefer answers the student must produce, such as a number typed into a [fill in the blank](https://edprocto.com/templates/fill_blank): a typed wrong answer often shows the method behind it.

## Should the answers show at once or wait for the lesson?

Either is defensible, because the research on feedback timing picks no winner. Kulik and Kulik (1988), pooling 53 studies, found that applied studies with real classroom quizzes usually favoured immediate feedback, while experiments on learning test content usually favoured a delay. In Mullet et al. (2014), engineering students whose homework feedback came a week after the deadline did better on later exams with new problems, yet believed immediate feedback had helped them more. Our guide to [study mode](https://edprocto.com/guides/study-mode-a-quiz-that-explains-each-answer-as-you-go) has the wider evidence.

Choose by what you want to see: EdProcto sets homework either way.

| | An ordinary quiz | A study quiz, shared by its code |
| --- | --- | --- |
| Marked | When it is handed in | Each answer as it is given |
| The student sees | The result at hand-in, answers shown or held as you set | The answer and its explanation after each question |
| You have | The result in your dashboard | The list of misses, the retry and the due date, which stay with your account |

To hold the answers in an ordinary quiz, set **Show the correct answer** and **Show the explanation** to "Never". With **Show the score** at "When released", the score stays hidden until you press "Release the scores" on the results page; at "Never" it is not shown at all.

Our rule, not a finding: hold the answers to hear the student's reasoning first, and show them when a wrong method would otherwise be practised all week.

## How do you read the results in the five minutes before the lesson?

Read the misses and the times, not the score. For one student's quiz, the average on EdProcto's results page is that student's score. On every plan the page says so itself, under "Read this first": "Fewer than five submissions: the numbers below describe these students, not the class or the teaching yet." On the Educator plan, the column for how well a question separates students shows a dash below six submissions.

1. **Close the quiz.** You open and close it by hand ("Publish & share", "Close the quiz"): it has no set hours, and EdProcto sends no reminders.
2. **Grade what is waiting.** A written answer counts as zero until it is marked. Mark it yourself, or press "Ask Procto AI" on that answer and keep or change its mark.
3. **Read the topics.** Give each question a topic in the builder. On the Educator plan the page lists the score on each, lowest first; on Free and Tutor, read "Got it" for the questions that share a topic.
4. **Read the time on each miss.** Slow and wrong is a student who tried. Fast and wrong did not see the difficulty.
5. **Open the submission.** It shows what the student gave for every question.

### One student, one week: a worked example

A student of fourteen had a lesson on percentage change that ended with reverse percentages. The homework was six fill in the blank questions: four on the lesson, one on ratio from four weeks back, and one a step ahead. Every figure was computed in code, and the result lines are what EdProcto's marking and analysis return for these answers.

| Question | Right | Typed | Time |
| --- | --- | --- | --- |
| Increase 240 by 15% | 276 | 276 | 0:35 |
| Decrease 350 by 8% | 322 | 322 | 0:40 |
| £80 rises to £92: the percentage increase | 15 | 15 | 0:55 |
| £68 after a 20% reduction: the price before | 85 | 81.60 | 1:35 |
| The smaller share of £120 split 3:5 | 45 | 45 | 0:50 |
| £200 rises by 10%, then 10% again | 242 | 240 | 0:30 |

The page reads: score 66.7%, time 5:05, and, in the Educator plan's full analysis, "Reteach before moving on: Reverse percentages (0%), Repeated change (0%)." Ignore the line saying the student passed: the default pass mark of 60% means nothing here, and a score is not a grade.

£81.60 is 68 × 1.2: the student added 20% of the sale price back on, and spent longest on it. That is a wrong idea, held with effort. £240 took thirty seconds: 10% of the starting amount, twice, on a question not yet taught. That is the way into this week's lesson.

The first ten minutes, planned from that:

- **Minute 1.** Say what held: percentage change three times out of three, and ratio a month on.
- **Minutes 2 to 5.** "Talk me through £81.60." Then: £68 is 80% of the old price, so divide by 0.8 to get £85. Two fresh ones follow: £46 after a 15% rise (£40) and £56 after a 30% cut (£80).
- **Minutes 6 to 8.** "What is 10% of £220?" The student finds £22, reaches £242, and today's topic begins from their own answer.
- **Minutes 9 and 10.** Agree the next quiz, where a reverse percentage returns in new numbers as the older question.

## What do you do when the homework is not done?

Do the quiz in the first five minutes of the lesson, without comment, then ask what got in the way. The student learns that the quiz happens either way.

The research points at follow-up, not penalties. Rosário et al. (2015) assigned teachers of English as a foreign language at random to one of five ways of following up homework for six weeks, and 26 of the 45, with 553 sixth graders, completed the study. Checking it orally, on the board, or by collecting and grading it led to better exam results than only checking that it was done or answering questions about it. The authors call the effects small.

So act on the content every week. EdProcto's results page counts an attempt begun and never handed in as in progress, which separates "forgot" from "got stuck". After two misses in a row, shrink the quiz, move the day and tell the parent: steps of our own.

## What do you tell the parent?

Tell the parent when the quiz is due, how long it takes, that it is done alone and that its score is not a grade. Then ask for one thing: a fixed time and place.

Patall et al. (2008) found, across 14 studies, that training parents to be involved in homework led to higher completion and fewer homework problems. Across 20 others, parent involvement went with lower achievement for middle school students and in maths, and rule-setting had a stronger link with achievement than other kinds of involvement. Help with the answers also spoils the reading: you would be planning from the parent's knowledge. The parent's side is in [checking a child's homework](https://edprocto.com/guides/how-to-check-your-childs-homework-with-a-quick-quiz).

Report in topics, not percentages. EdProcto sends no result email to a student or a parent, so the report is a message you write.

## How does EdProcto set homework for tutoring students?

EdProcto is the best tool for homework between lessons because it does the part the research credits: it puts the result in front of the tutor before the lesson, question by question, with the time spent on each. In the Maine trial the gain came with teachers who were given that information and trained.

**Writing it.** Paste the lesson's notes into the quiz maker on the [home page](https://edprocto.com/) and pick a shape: Procto AI drafts the questions from your text, and you edit them. The [kits](https://edprocto.com/kits) hold ready-made quizzes by subject and level.

**Sending it.** Publish, and send the link or the code. With "An account is needed to take it" on, the student types a name and an email address and needs no account of their own. Leave the security off or light: a logged event is a record for you, never proof, and no setting makes cheating impossible ([integrity](https://edprocto.com/integrity)).

**Across the weeks.** The Students page keeps one line a student, matched by the email address typed: quizzes handed in, their average, security events and the last quiz with its date. It does not chart a topic from quiz to quiz, so that story is read from each quiz's own results.

**What it costs.** The Free plan's ten quizzes a month cover a weekly quiz for two students, marked by rule. The Tutor plan removes the limit, opens all 111 templates, has Procto AI write 50 questions a month from a PDF, a Word file or your notes, and adds its grading of open answers. The Educator plan adds certificates and the full analysis: the score on each topic and the line on what to reteach. See [pricing](https://edprocto.com/pricing).

## Quick questions

### How much homework should a tutor give?

One short quiz a week: about four questions under eleven, six to eight for a teenager, ten over sixteen. That is our working rule, not a research finding.

### How do I set homework as a tutor without marking it by hand?

Set an online quiz with typed or chosen answers, which are marked by rule at hand-in. EdProcto's page [for tutors](https://edprocto.com/solutions/tutors) shows the loop. Written answers still wait for you.

### What if the student gets help with the quiz?

Then the result describes the helper. Ask the parent for a time and a place, not for help, and ask one question again on paper if a score looks too good.

### Is online homework better than homework on paper?

Not by itself: in a review of 31 studies, half found no difference between the formats. The large school trial that found a gain also gave teachers the results and training.

## References

- Cooper, H., Robinson, J. C., & Patall, E. A. (2006). Does homework improve academic achievement? A synthesis of research, 1987–2003. *Review of Educational Research, 76*(1), 1–62. [https://doi.org/10.3102/00346543076001001](https://doi.org/10.3102/00346543076001001)
- Fan, H., Xu, J., Cai, Z., He, J., & Fan, X. (2017). Homework and students' achievement in math and science: A 30-year meta-analysis, 1986–2015. *Educational Research Review, 20*, 35–54. [https://doi.org/10.1016/j.edurev.2016.11.003](https://doi.org/10.1016/j.edurev.2016.11.003)
- Fernández-Alonso, R., Suárez-Álvarez, J., & Muñiz, J. (2015). Adolescents' homework performance in mathematics and science: Personal factors and teaching practices. *Journal of Educational Psychology, 107*(4), 1075–1085. [https://doi.org/10.1037/edu0000032](https://doi.org/10.1037/edu0000032)
- Kulik, J. A., & Kulik, C.-L. C. (1988). Timing of feedback and verbal learning. *Review of Educational Research, 58*(1), 79–97. [https://doi.org/10.3102/00346543058001079](https://doi.org/10.3102/00346543058001079)
- Magalhães, P., Ferreira, D., Cunha, J., & Rosário, P. (2020). Online vs traditional homework: A systematic review on the benefits to students' performance. *Computers & Education, 152*, Article 103869. [https://doi.org/10.1016/j.compedu.2020.103869](https://doi.org/10.1016/j.compedu.2020.103869)
- Mullet, H. G., Butler, A. C., Verdin, B., von Borries, R., & Marsh, E. J. (2014). Delaying feedback promotes transfer of knowledge despite student preferences to receive feedback immediately. *Journal of Applied Research in Memory and Cognition, 3*(3), 222–229. [https://doi.org/10.1016/j.jarmac.2014.05.001](https://doi.org/10.1016/j.jarmac.2014.05.001)
- Patall, E. A., Cooper, H., & Robinson, J. C. (2008). Parent involvement in homework: A research synthesis. *Review of Educational Research, 78*(4), 1039–1101. [https://doi.org/10.3102/0034654308325185](https://doi.org/10.3102/0034654308325185)
- Rawson, K. A., & Dunlosky, J. (2022). Successive relearning: An underexplored but potent technique for obtaining and maintaining knowledge. *Current Directions in Psychological Science, 31*(4), 362–368. [https://doi.org/10.1177/09637214221100484](https://doi.org/10.1177/09637214221100484)
- Roediger, H. L., & Karpicke, J. D. (2006). Test-enhanced learning: Taking memory tests improves long-term retention. *Psychological Science, 17*(3), 249–255. [https://doi.org/10.1111/j.1467-9280.2006.01693.x](https://doi.org/10.1111/j.1467-9280.2006.01693.x)
- Rosário, P., Núñez, J. C., Vallejo, G., Cunha, J., Nunes, T., Suárez, N., Fuentes, S., & Moreira, T. (2015). The effects of teachers' homework follow-up practices on students' EFL performance: A randomized-group design. *Frontiers in Psychology, 6*, Article 1528. [https://doi.org/10.3389/fpsyg.2015.01528](https://doi.org/10.3389/fpsyg.2015.01528)
- Roschelle, J., Feng, M., Murphy, R. F., & Mason, C. A. (2016). Online mathematics homework increases student achievement. *AERA Open, 2*(4). [https://doi.org/10.1177/2332858416673968](https://doi.org/10.1177/2332858416673968)
- Trautwein, U. (2007). The homework–achievement relation reconsidered: Differentiating homework time, homework frequency, and homework effort. *Learning and Instruction, 17*(3), 372–388. [https://doi.org/10.1016/j.learninstruc.2007.02.009](https://doi.org/10.1016/j.learninstruc.2007.02.009)