A quiz written for a ten-year-old and shrunk for a five-year-old tests the wrong things: reading the stem, aiming at a small button and sitting through twenty items. A child who can count nine ducks scores nothing, and the record says "cannot count". Quiz questions for young children have to be built the other way round, from what a small hand and an early reader can do.
EdProcto has a family of shapes made for that, led by big buttons: multiple choice with targets a small hand can hit. This guide is for early years teachers and for parents.
The short version: one tap to answer, a spoken prompt or a picture, two or three options, about five questions, right or wrong at once with the reason, and new pictures every time.
What can a child of four to seven do on a touch screen?
A child of four to seven can tap a large target almost every time, can usually drag one thing a short way with one finger, and cannot be relied on for a double tap, a long press or anything two-handed.
The fullest study is Vatavu et al. (2015): 89 children aged 3 to 6, tested on a phone and a tablet.
| Gesture | Attempts completed |
|---|---|
| Tap | 98.7% |
| Drag and drop with one finger | 92.0% |
| Double tap | 82.8% |
| Drag two targets at once, both hands | 53.7% |
Age mattered, but not for the tap. Across the four tasks the under-fours completed 73.5% of attempts and the over-fives 89.2%, yet every age group tapped about equally well. Only 24 of the 89 had used a touch screen before, so a child who uses one daily may beat these figures.
Tapping against dragging
Build every question you can on a single tap, the one gesture that works at every age here. Abdul Aziz et al. (2014) tested children aged 2 to 4 on an iPad, in the United Kingdom and again in Malaysia: four-year-olds managed all seven common gestures, while many two-year-olds struggled with drag and drop, pinch, spread and free rotation. Their summary gives no head count.
Dragging costs something even when it works: in the Vatavu study a one-finger drag failed on 11.5% of tablet attempts against 4.2% on the phone. Keep drags short, and prefer a shape where a tap does the same work: in spelling with letter tiles a tapped tile goes to the first empty place.
How big, and how far apart?
Make each target far bigger than an adult needs, with clear space round it. In the same study a three-year-old's tap landed on average 4.5 mm from the target's centre and an over-five's 3.4 mm, and the authors recommend that a screen for three-year-olds accept taps up to 10 mm off. Some under-fours held a tap for as long as 5.1 seconds, so a press must count however long it lasts.
Older children still miss. Anthony et al. (2012) gave 16 children aged 7 to 16 and 14 adults a tapping task on a phone. The children's first touch missed 23.1% of the time, the adults' 16.9%. When the screen moved on after a good tap, people tapped again on the old spot, and 96% of those stray taps were children's. No five-year-olds took part, so read it as a warning about layout: do not let a question advance the instant it is answered.
Our guide to hotspot questions sets out the accessibility standard's sizes for pointer targets. Treat those as the least a young child needs.
How little reading should a question need?
As near to none as you can manage: a spoken prompt, a picture, or one word the class has been taught. If a child must read a sentence to learn what is being asked, the quiz measures reading first.
A spoken prompt is realistic from about three and a half. Hiniker et al. (2015) watched 34 children aged 2 to 5 respond to the prompts common in touch-screen apps for pre-schoolers. Children under 3 followed an instruction only when an adult modelled it. By 3 and a half they could follow spoken instructions from the app and demonstrations on the screen. Prompting with a visual change alone did not work at any age studied.
- Say it. One short sentence in a clear voice: "Which one is the apple?" EdProcto's listen and point pairs a spoken prompt with pictures, and match the sound plays a clip with pictures as the options.
- Show it. The picture is the question: nine ducks, and numbers to tap. One taught word on a button at most.
- Read it aloud yourself where there is no recording: that helps with the screen, not the answer.
How many options, how many questions, and how long?
Two or three options, about five questions, and a few minutes with no clock. These are working rules, not findings: we found no study that measured the best number of questions at four to seven, or a minimum button size for quiz options.
- Options. Each is a picture to study and a target to hit. Three good options are enough even for adults, as the guide to multiple-choice questions explains, and two is a fair start.
- Questions. About five at four or five, rising to eight by seven, one idea to each. The ready-made early years kits are mostly ten questions long: with a four-year-old, treat five as a full sitting.
- Time. Three to five minutes, with no timer. Stop when the child has had enough.
What to do about guessing with two options
Do not judge from one sitting. Guessing alone gets four or five right out of five 19% of the time with two options and 5% with three (our sums).
- Move to three options once you can find two honest wrong ones.
- Read several sittings together. Twelve or more right out of fifteen two-option questions happens by luck less than 2% of the time.
- Move the right answer about, so a child who always taps one side is not right by habit.
- Ask how they knew. "How did you know it was four?" tells you more than the tap.
Never take marks away for a wrong tap.
What feedback does a small child understand?
A small child understands feedback that comes at once, says right or wrong plainly, and shows or says why. A total at the end means little to a five-year-old, and one set beside another child's teaches nothing.
Most apps stop at right or wrong. Tärning (2018) reviewed 242 apps used in Swedish primary schools: 77% said whether the answer was correct, only 12% explained why a wrong answer was wrong, and none of the encouragement they gave was about the task or the effort.
- At once, with the reason shown or said. "There are four. One, two, three, four", with a finger on each duck.
- About the task. "You counted every duck", not "clever girl". The ten-minute routine for parents covers the research on praise.
- Nothing to compare. No rank, no class average, no leaderboard.
Why change the pictures every time?
Because a young child remembers where the answer was sooner than why it was right. Give the same five pictures in the same order twice and the second result measures memory of the quiz: the duck question is the left button.
Keep the idea and change the surface: other objects, another number of them, the right answer somewhere new. Recalling a thing does strengthen memory, but experiments such as Karpicke et al. (2016) tested children of about ten, not four-year-olds, so the size of the effect may not carry over.
EdProcto's early years kits are made by rule for this reason: Counting to 10, for ages 4 to 5, draws fresh sets every time it is opened.
Which shape fits which skill?
Choose the shape by what the child must do to show the skill, and prefer one answered with a tap.
| Skill | Shape in EdProcto | Its writing note |
|---|---|---|
| Pointing at part of a picture | Touch the picture | The zone is the whole part, not its centre |
| Counting objects | Count the objects | Objects in rows for the youngest |
| Hearing a word, choosing its picture | Listen and point | One clear voice, three pictures at most |
| Matching pictures | Picture match | Four pairs at most |
| Spelling with letter tiles | Spell it | One extra tile, so the last letter is not free |
| Continuing a pattern | Build the pattern | One rule, shown at least twice |
| Any choice among two to four | Big buttons | Big and far apart |
All seven are ready in the builder and marked by rule, and so is a path to trace with a finger. Make wrong options wrong in idea, not confusing to the eye: for "apple", a pear and a banana, not a red ball.
A five-question quiz, worked
For five-year-olds after a week on the farm, with every prompt spoken:
- Count the objects. Four ducks in a row. "How many ducks?" A pad of large digits: the child taps 4.
- Listen and point. "Which one is the cow?" Pictures: cow, horse, sheep.
- Touch the picture. "Touch the dog's tail." The zone covers the whole tail.
- Build the pattern. Egg, chick, egg, chick, egg. "What comes next?"
- Big buttons. "Which is bigger?" Two buttons: 7 and 4.
Each is one of EdProcto's young learner templates. Tomorrow's version keeps the five ideas and changes every picture and number.
What should you watch for while the child takes it?
Watch the hands and the face, not the score: they tell you whether a wrong answer came from the knowledge or from the screen.
- A tap just outside the button. The screen failed, not the child. Make the targets bigger.
- Extra taps after the answer registered. Check they have not answered the next question.
- A palm, a knuckle or two fingers on the glass. In Vatavu et al. (2015) something other than the index finger was used in 17% of cases on the tablet.
- A look at you before each tap. Keep your face neutral.
Some children in that study stopped unless the adult beside them encouraged them, and one apologised to the device after an error. Stay close.
What are the common mistakes?
The common mistakes all make the quiz measure something other than the knowledge.
- A shrunk exam: small buttons and a sentence to read.
- Wrong options that only look alike.
- Double taps, long presses and long drags.
- A timer, a total or a rank on the child's screen.
- One sitting read as a verdict.
What can a quiz like this not tell you?
A quiz like this cannot diagnose anything, and it is not a developmental screen. It tells you that today, on this screen, this child tapped the right picture for these five questions. That is enough to choose what to practise next, and no more.
Proper assessments on tablets exist, and they are a different kind of thing. Pitchford and Outhwaite (2016) built one for the first years of primary school, then tested 283 children in Malawi and 70 in the United Kingdom to check its reliability and validity, and still call the result a proof of concept.
- A low score says the questions, the screen or the day did not suit. It says nothing about what a child is able to learn.
- A worry about hearing, sight, speech, movement or development belongs with the child's teacher or a health professional. A quiz result is not evidence either way.
How does EdProcto build and mark quiz questions for young children?
EdProcto is the best tool for a quiz a non-reader can take, because most of its young learner shapes are answered with a single tap, the one gesture nearly every child in the Vatavu study could make.
The seven ready shapes are marked by rule, and in study mode the answer and a plain explanation follow straight away. "Show the score" can be set to Never, and the class average stays off unless you switch it on.
The early years kits each name their ages, such as Counting to 10 (4 to 5) and Shapes (4 to 6). Opposites and First sounds (both 4 to 6) show words, so read them aloud until the child can. The parent's dashboard then lists what was right, what was missed and what to try again. For a child who has begun to read, paste a page of the reading book into the quiz maker on the home page and Procto AI drafts the questions.
The rule-made drills are free without limit. Free also gives ten quizzes a month with the core templates, Tutor opens all 111 templates, and Educator adds the full analysis. See pricing.
Quick questions
At what age can a child take a quiz on a tablet?
From about four, if the answer is one tap on a large target and the question is spoken or shown as a picture. In Vatavu et al. (2015), children aged 3 to 6 completed 98.7% of their taps.
How many questions should a quiz for a five-year-old have?
About five, in three to five minutes, with no timer. That is a working rule: we found no study that measured it.
Should a quiz for young children show a score?
Not one the child can compare. Show right or wrong and the reason after each question. In study mode EdProcto's big buttons shows both at once, and the score can be hidden.
Can a quiz show whether my child is behind?
No. It shows what was tapped today on a few questions. It is not a diagnosis or a developmental screen, and a low score is a reason to change the quiz.
References
- Abdul Aziz, N. A., Mat Sin, N. S., Batmaz, F., Stone, R., & Chung, P. W. H. (2014). Selection of touch gestures for children's applications: Repeated experiment to increase reliability. International Journal of Advanced Computer Science and Applications, 5(4). https://doi.org/10.14569/IJACSA.2014.050415
- Anthony, L., Brown, Q., Nias, J., Tate, B., & Mohan, S. (2012). Interaction and recognition challenges in interpreting children's touch and gesture input on mobile devices. In Proceedings of the 2012 ACM International Conference on Interactive Tabletops and Surfaces (pp. 225–234). Association for Computing Machinery. https://doi.org/10.1145/2396636.2396671
- Hiniker, A., Sobel, K., Hong, S. R., Suh, H., Irish, I., Kim, D., & Kientz, J. A. (2015). Touchscreen prompts for preschoolers: Designing developmentally appropriate techniques for teaching young children to perform gestures. In Proceedings of the 14th International Conference on Interaction Design and Children (pp. 109–118). Association for Computing Machinery. https://doi.org/10.1145/2771839.2771851
- Karpicke, J. D., Blunt, J. R., & Smith, M. A. (2016). Retrieval-based learning: Positive effects of retrieval practice in elementary school children. Frontiers in Psychology, 7, Article 350. https://doi.org/10.3389/fpsyg.2016.00350
- Pitchford, N. J., & Outhwaite, L. A. (2016). Can touch screen tablets be used to assess cognitive and motor skills in early years primary school children? A cross-cultural study. Frontiers in Psychology, 7, Article 1666. https://doi.org/10.3389/fpsyg.2016.01666
- Tärning, B. (2018). Review of feedback in digital applications: Does the feedback they provide support learning? Journal of Information Technology Education: Research, 17, 247–283. https://doi.org/10.28945/4104
- Vatavu, R.-D., Cramariuc, G., & Schipor, D. M. (2015). Touch interaction for children aged 3 to 6 years: Experimental findings and relationship to motor skills. International Journal of Human-Computer Studies, 74, 54–76. https://doi.org/10.1016/j.ijhcs.2014.10.007