# Picture matching quizzes for young children: what a match shows

**How do I make a picture matching quiz for young children, and what does a finished match show?** A picture matching quiz for young children of four to seven sets three to five pairs, picture to picture or picture to word, joined by tapping a card, with no typing. Keep one kind of link in each set, use clear pictures the child can name, treat the last pair as unproven, and read which pairs were swapped.

Published 2026-10-10 by EdProcto. https://edprocto.com/guides/picture-matching-quizzes-for-young-children

Matching activities are among the first a child can do with a finger and no pencil: this one goes with that one. Picture matching quizzes for young children give a child of four to seven a few pairs to join, picture to picture or picture to word, and what the finished set shows depends on how the pairs were chosen.

This guide is published by EdProcto, whose shape for the task is [Picture match](https://edprocto.com/templates/picture_match). You type a label for each card and, where you have one, a picture address to go with it; the child sees the pictures on large cards and taps the match, and every pair is marked by rule. The guide is for an early years teacher, a teaching assistant or a [parent](https://edprocto.com/solutions/parents).

The studies were read on 10 October 2026; those read only as abstracts are marked.

## What does a matching task ask of a young child?

A matching task asks three things: seeing what two things share, holding one of them in mind while searching for the other, and, where one side is print, a first reading of a whole word.

**Seeing what is shared.** Young children pair by look before anything else. In Diesendruck and Bloom (2003; abstract read), three-year-olds went by shape, over size, colour or material, when asked to choose an object of the same kind as another.

**Holding one in mind.** The child looks at the dog, keeps it, and searches the far side for the kennel. That store, working memory, is small, and its measured size depends on the task. Simmering and Perone (2013) gathered estimates from 50 papers. In two studies that flashed up coloured squares and asked whether one had changed, a five-year-old's estimate was 1.5 items in one and 2.8 in the other, and a seven-year-old's 2.9 and 3.8; the two worked the figure out differently.

### The first reading of a whole word

A picture-to-word match is often a child's first reading of whole words, and it can be got right without reading the whole word. Bowman and Treiman (2008; abstract read) taught made-up spellings to children of about four years nine months who could not yet read: those in which a letter's name is heard in the word (AP for ape) were learned more easily than arbitrary ones (OM for ape). A right match may rest on one letter.

A picture beside a word also does the reading for the child. Hill (1995) taught twelve short words to several groups, one of them ten typically developing children aged four and five: four words alone, four under a line drawing and four under a colour picture. Over six rounds these children said the pictured words every time, 24 of 24. Shown the same words alone, they read 4.3 of 24 for line drawings and 3.2 for colour pictures, against 8.4 for words taught alone. The study was small and exploratory. For matching, it argues for counting a word as read once it is read with its picture out of sight.

## Which kinds of match are there, and what can each show?

There are five kinds, and each shows something different.

| Kind of match | A pair | What a right pair can show | In EdProcto |
| --- | --- | --- | --- |
| The same thing | Two cats in different poses | Seeing past colour, size and pose | [Picture match](https://edprocto.com/templates/picture_match), a picture on each card |
| Things that go together | Bird and nest | Knowledge of the world | Picture match |
| A thing and its group | Apple and "fruit" | A category held as a rule | [Sort into groups](https://edprocto.com/templates/categorize) |
| A thing and its word | A dog and "dog" | The beginning of word reading | Picture match, words on the right |
| A thing and its first sound | A sun and "s" | Hearing the first sound of a spoken word | Picture match, letters on the right |

Keep one kind to a set, and name it in the instruction. Nguyen and Murphy (2003) showed children a food and asked which of two others was "the same kind of food". Set against an unrelated food, the related one was picked 70% of the time by 16 four-year-olds and 85% by 16 seven-year-olds. When one choice shared the category (another fruit) and the other the occasion (another breakfast food), the four-year-olds picked the category 44% of the time, which is not different from chance, and the seven-year-olds 62%. Both links are real to a young child, so a set that mixes them has two defensible answers. "Where does each animal live?" names the link. "Match them up" leaves it open.

A thing and its group is really a sort, since several things share one partner: see [sorting questions](https://edprocto.com/guides/sorting-questions-drag-items-into-categories). Rhymes and first sounds are in [quizzes for children who cannot read yet](https://edprocto.com/guides/quizzes-for-children-who-cannot-read-yet-pictures-and-sounds).

## How many pairs suit which age, and why does the last pair give itself away?

Three pairs at four and five, four at five and six, five by seven; and the last pair gives itself away because only one partner is left. The numbers are working rules: no study we read tested the number of pairs in a set at this age. EdProcto's own tip for the shape is "Four pairs at most for the youngest."

Four pairs put eight pictures in front of a child who holds two or three in mind. Past five pairs the task turns into searching.

### The last pair proves nothing

When each partner is used once, the last pair needs no knowledge, and children can work that out from about three. Mody and Carey (2016) hid a sticker in one cup of each of two pairs and showed one cup to be empty, so its partner had to hold a sticker. Choosing among the three cups left, 24 children at each age picked that partner 36% of the time at two and a half, about what chance gives, 58% at three, 64% at four and 76% at five.

A one-to-one set has an odd arithmetic too: a child can never get exactly one pair wrong. Four pairs come out with four, two, one or none right (our sums), so "all four" may be three known and one left over. Two habits keep the result honest:

1. On cards, add one spare picture to the answer side.
2. Ask about the last pair aloud: "How did you know the spider goes there?"

## How do you choose pictures a child reads the way you meant?

Choose pictures the child names the way you do, and check by asking. Wang et al. (2014) had 32 Mandarin-speaking children of about four and 34 of about six name 260 standard line drawings. The four-year-olds' commonest name was not the expected one for 88 of the 260 pictures, and the six-year-olds' for 73. More than half the time the four-year-olds gave a broader word (clothes for a blouse, a jacket or a coat), and sometimes a near relation (apple for a peach).

1. **Write the instruction first.** It fixes the link.
2. **One object to a picture,** whole, in clear outline, on a plain ground, all in one style and size.
3. **Familiar things.** A nest from this week's story, not a burrow nobody has met.
4. **No look-alikes, unless looking closely is the point.** An orange beside a ball tests shape.
5. **A naming round.** Before the set counts, point to each picture and ask "What is this?" Use the child's word, or change the picture.
6. **Words in one print and length.** Once first letters are doing the work, set two words that start alike, such as dog and duck.
7. **Type the pairs into EdProcto's [Picture match](https://edprocto.com/templates/picture_match)** and try the set on the child's own device.

## Should a child tap, drag or point?

Use whichever costs the hand least, because the gesture is not what is learned. Russo-Johnson et al. (2017) had 170 children aged two to four learn four new object names in an app by watching, tapping or dragging the named object. No way of interacting was better overall: what differences there were depended on the child.

Dragging also goes wrong more often when the target is small. Cassidy et al. (2019; abstract read, which gives no ages) measured 28 children dragging and dropping on a tablet: the error rate was 12.6% with a finger against 9.3% with a stylus, and it roughly doubled with smaller targets. On a laptop, let the child point at the screen while an adult works the pointer. Sizes for touch targets are in [questions for young learners](https://edprocto.com/guides/questions-for-young-learners-big-targets-little-reading).

## How do you read the results, and what do you teach next?

Read which pairs were swapped with which, because the total hides it. Note the pairs as the child makes them, or read them back afterwards.

| What you see | What it suggests | What to teach next |
| --- | --- | --- |
| Two partners swapped | The two are confused with each other | Set the two side by side and say what differs |
| One partner chosen several times | The other partners are unknown | Teach those from real things or a book |
| All right, the last by default | One pair unproven | Ask about it, or begin with it next time |
| Wrong on a picture named differently | The picture failed | Change the picture |

One sitting is thin evidence: three pairs joined blindly come out all right one time in six (our sums). A set that goes wrong tells you what to practise next. It is no verdict on the child.

## How often, and for how long?

A few minutes at a time, on several days. No study we read timed matching sets, so the rule is borrowed from lessons. Vlach and Sandhofer (2012) gave 36 children aged five to seven four science lessons of about five minutes each: all in one day, two a day, or one a day. A week after the last lesson, the one-a-day group scored 3.08 of 4 on the harder idea, against 2.08 for two a day and 1.75 for all in one day. Come back tomorrow with the same link and different pictures.

## A worked example: four animals and their homes

The two children are invented for this example.

The instruction, said aloud: "Where does each animal live?" Eight cards, each a black outline on white:

- **Animals:** a bird, side on; a striped bee with two wings; a dog, standing; a spider from above, eight legs showing.
- **Homes:** a nest, a bowl of twigs holding three eggs; a hive, a domed straw basket with a small doorway; a kennel, a hut with an arched door; a web, a wheel of threads between two twigs.

In the naming round both children call the hive "a basket", so the adult names it once.

- **Ava, five:** bird and nest, dog and kennel, bee and web, spider and hive. Two of four.
- **Leo, five:** bird and nest, dog and kennel, bee and nest, spider and nest. Two of four.

The same total, and two different lessons. Ava swapped the hive and the web: she knows both are homes for small creatures and not whose is whose, so the next step is those two side by side. Leo went back to the nest twice: the hive and the web are both new to him, and a picture book about minibeasts comes before another set.

## How does EdProcto build and mark a picture matching quiz?

EdProcto is the best tool for keeping the record of a matching set, because it marks each pair by rule and keeps every partner on the cards to the end, so the screen gives nothing away: in Mody and Carey (2016) three-year-olds already used elimination once one choice was ruled out.

**Building.** In the builder a Picture match question has a "Prompt" box, whose example reads "Match each animal to its home.", and a list headed "Pairs". Each row carries a "Left label" and a "Right label", and under them a "Left picture address" and a "Right picture address" you fill where you have a picture; a fresh row waits under the last pair. The label is the key the pair is marked against and the name a screen reader reads; the picture is what the child sees on the card. A label can be a word, a letter, a number, or a symbol such as △, which is how the demonstration on the template's page sets a shape against its name. Give every left-hand label its own wording. Set "Points" to the number of pairs.

**Answering.** The child sees each left-hand card with its picture, and under it the right-hand items stand as large buttons, each a picture, a word, or both, in a shuffled order. Tapping a button joins the pair; a second tap keeps it, and tapping another button changes the choice. No typing, and on a laptop the child can point while an adult taps.

**Marking.** With "Partial credit" on, the mark is the share of pairs that are right. In EdProcto's marking function a four-pair set worth four points gave 2 for Ava's swap, 2 for Leo's answer, 3 for three right with one line left at "Choose", and 1 when one partner was picked on every line. With it off, each of those answers gave 0: only the complete set earns the points.

**Afterwards.** Choose the study kind of quiz, which the builder describes as "Each answer shown at once, misses retried." The child's result lists "Your answer:" under each question, where the swaps are read back, and the results page gives "Got it" for each question on every plan. "Show the score" can be set to "Never".

**Real pictures on the cards.** Picture match takes a picture address beside each label, so the eight cards can carry the animals and their homes themselves. [Pick the picture](https://edprocto.com/templates/pick_image) and [Listen and point](https://edprocto.com/templates/listen_point) take an "Image address" for each option too, where you would rather ask the same four pairs as four picture questions: "Which one is the bee's home?"

**Plans.** On the Free plan a teacher makes ten quizzes in a month from the core templates, with marking by rule and up to thirty children on each. The Tutor plan carries every one of the 111 [templates](https://edprocto.com/templates), and the full analysis comes with Educator: [pricing](https://edprocto.com/pricing) has the detail.

## Quick questions

### What is a picture matching quiz for young children?

A short set of pairs, picture to picture or picture to word, joined by tapping a card. EdProcto's [Picture match](https://edprocto.com/templates/picture_match) marks each pair by rule.

### How many pairs should a matching set have for a five-year-old?

Three or four, as a working rule. EdProcto's own tip for the shape is four at most for the youngest.

### Why is the last pair in a matching set free?

Because one partner is left. In Mody and Carey (2016) children from three upwards chose by elimination more often than chance, so ask how the child knew.

### What does a part-right matching set earn in EdProcto?

With "Partial credit" on, the share of pairs that are right: two of four pairs earn half the points. With it off, a set earns its points only when every pair is right.

## References

- Bowman, M., & Treiman, R. (2008). Are young children logographic readers and spellers? *Scientific Studies of Reading, 12*(2), 153–170. [https://profiles.wustl.edu/en/publications/are-young-children-logographic-readers-and-spellers/](https://profiles.wustl.edu/en/publications/are-young-children-logographic-readers-and-spellers/)
- Cassidy, B., Read, J. C., & MacKenzie, I. S. (2019). FittsFarm: Comparing children's drag-and-drop performance using finger and stylus input on tablets. In *Human-computer interaction: INTERACT 2019* (Lecture Notes in Computer Science, Vol. 11748, pp. 656–668). Springer. [https://knowledge.lancashire.ac.uk/id/eprint/30143/](https://knowledge.lancashire.ac.uk/id/eprint/30143/)
- Diesendruck, G., & Bloom, P. (2003). How specific is the shape bias? *Child Development, 74*(1), 168–178. [https://doi.org/10.1111/1467-8624.00528](https://doi.org/10.1111/1467-8624.00528)
- Hill, L. (1995). An exploratory study to investigate different methods for teaching sight vocabulary to people with learning disabilities of different aetiologies. *Down Syndrome Research and Practice, 3*(1), 23–28. [https://doi.org/10.3104/reports.48](https://doi.org/10.3104/reports.48)
- Mody, S., & Carey, S. (2016). The emergence of reasoning by the disjunctive syllogism in early childhood. *Cognition, 154*, 40–48. [https://pmc.ncbi.nlm.nih.gov/articles/PMC4939102/](https://pmc.ncbi.nlm.nih.gov/articles/PMC4939102/)
- Nguyen, S. P., & Murphy, G. L. (2003). An apple is more than just a fruit: Cross-classification in children's concepts. *Child Development, 74*(6), 1783–1806. [https://doi.org/10.1046/j.1467-8624.2003.00638.x](https://doi.org/10.1046/j.1467-8624.2003.00638.x)
- Russo-Johnson, C., Troseth, G., Duncan, C., & Mesghina, A. (2017). All tapped out: Touchscreen interactivity and young children's word learning. *Frontiers in Psychology, 8*, Article 578. [https://doi.org/10.3389/fpsyg.2017.00578](https://doi.org/10.3389/fpsyg.2017.00578)
- Simmering, V. R., & Perone, S. (2013). Working memory capacity as a dynamic process. *Frontiers in Psychology, 3*, Article 567. [https://doi.org/10.3389/fpsyg.2012.00567](https://doi.org/10.3389/fpsyg.2012.00567)
- Vlach, H. A., & Sandhofer, C. M. (2012). Distributing learning over time: The spacing effect in children's acquisition and generalization of science concepts. *Child Development, 83*(4), 1137–1144. [https://doi.org/10.1111/j.1467-8624.2012.01781.x](https://doi.org/10.1111/j.1467-8624.2012.01781.x)
- Wang, L., Chen, C.-W., & Zhu, L. (2014). Picture norms for Chinese preschool children: Name agreement, familiarity, and visual complexity. *PLoS ONE, 9*(3), Article e90450. [https://doi.org/10.1371/journal.pone.0090450](https://doi.org/10.1371/journal.pone.0090450)