Hughes policeman doll study (1975) tested Piaget’s claim that young children are egocentric, unable to see another person’s point of view. Three- and four-year-olds hid a boy doll from toy policemen. Most hid him where no policeman could see him, even from two or three at once. Hughes and Donaldson concluded that Piaget had overstated young children’s egocentrism.
Key Takeaways
- Aim: To test Piaget’s claim that young children cannot take another person’s point of view into account. Hughes used a hiding game with toy policemen, not the three mountains task.
- Method: Three studies used different children, 20, 30 and 40 three- and four-year-olds. The paper does not say where or how they were recruited. Each child hid a boy doll (in Study One, by placing a wall) so that no toy policeman could see him. Study One used one policeman and one wall in three positions, and every child did all three (repeated measures). Study Two used two policemen and four trials, after questions that checked the child understood the set-up. Study Three used two matched groups, one with five sections and two policemen and one with six sections and three policemen (independent groups). In Study Three the independent variables were the task version and age (a natural variable); for an exam, task version is the clearest. The dependent variable was whether the child hid the boy where no policeman could see him.
- Results: Study One was almost errorless, with nine of ten children in each age group placing the wall correctly every time. In Study Two, 27 of 30 children (90%) were right on at least three of four trials. In Study Three, 9 of 10 four-year-olds made no errors with two policemen and 8 of 10 with three. Three-year-olds made more errors. Their 6 of 10 and 7 of 10 made one error or none, a looser cut-off than the 9 and 8 of 10 with no errors. The authors’ test found the age difference too big to put down to chance with two policemen only. Three policemen were not harder than two. No child consistently hid the boy from their own view.
- Conclusion: In the hiding game, three- and four-year-olds could work out what the policemen could see (most three-year-olds and almost all four-year-olds). The authors said this added to evidence that Piaget overstated egocentrism. They suggest the game made sense to the children, but the studies did not test this. The study does not show that these children could pass the mountain task.
- Strengths: The results are consistent (an exam would call this reliable). The same finding recurred in three studies with different children and one, two or three policemen. The measure was clear, because the child either hid the boy or did not. Two children who played a different game were scored wrong, reluctantly.
- Limitations: The samples were small, with only 10 children per group in Studies One and Three, so the results may not apply to all children. The study also tests a different skill, because the policemen tasks ask whether someone can see something, not how a scene looks from another position, and the children did not do the mountain task.
- Ethical Issues: By modern standards, and although harm was probably minimal (the children obviously enjoyed the games), two issues arise. First, the paper does not mention a parent’s permission, so ask a parent first and explain what the child will be doing. Second, the paper does not mention the child’s right to stop, so tell each child they can stop. The paper also does not mention debriefing, so explain the study afterwards.
Aim
Piaget’s theory says that children below about seven are egocentric. He called this early period the preoperational stage.
That means they find it hard to see that other people’s view differs from their own. It is a limit on thinking, not selfishness.
Piaget and Inhelder tested this with a model of three mountains. A child sat in front of it. A doll was placed on another side of the model.
The child was shown pictures of the mountains from different angles. The child had to choose the picture that shows what the doll sees.
Young children tended to pick their own view (Piaget & Inhelder, 1956, as described by Hughes & Donaldson, 1979). This is the three mountains task.
Hughes and Donaldson argued that this is not the only way to test whether children can see other points of view. They suggested the child may not understand what to do.
So they used a hiding game with a boy doll and toy policemen. They designed it so that a child cannot succeed without working out what the policemen can see.
These aims can be put as three questions.
- One Policeman: Can three- and four-year-olds hide a boy doll from one toy policeman?
- Two Policemen: Can they hide the boy from two policemen at once, so that neither can see him?
- Harder Versions: Do they still succeed with more hiding places and three policemen, and does age matter?
Method
Design and Variables
Picture a low table with toy dolls and walls. In Study One the policeman and boy were about 6 cm tall and the wall 7 cm high.
The child’s job is to hide the boy so that no policeman can see him. A policeman sees the boy along a line of sight, the straight line between them. A wall on that line blocks the view.
In Studies Two and Three, the walls make sections, which are spaces where the boy can be put.
Hughes and Donaldson (1979) ran three studies, called Study One, Study Two and Study Three here. Each child had several goes, called trials.
Every experiment has two kinds of variable. The independent variable is what the researcher changes. The dependent variable is what the researcher measures.
Age is a natural variable. The researchers cannot change it, so they compare children who are already three or four. The paper does not use these terms, so the labels in the table are ours.
Each study below explains the positions and sections in the table.
| Study | Independent variable | Dependent variable |
|---|---|---|
| One: one policeman | The child’s age (natural variable) and the position of the policeman and boy (three positions). | Whether the child put the wall so the policeman could not see the boy. |
| Two: two policemen | Nothing was compared. Every child did the same four trials. | Whether the child hid the boy so neither policeman could see him, counted over the four trials. |
| Three: harder versions | The child’s age (natural variable) and the task version (two policemen and five sections, or three and six). | Whether the child hid the boy so no policeman could see him, counted as errors. |
The design words are ours too. In Study One each child did all three positions. Our label for this is repeated measures. Study Three used different children in each group. Our label for this is independent groups. Study Two had one group and no comparison group.
For an exam, take Study Three. The clearest pair is the task version (two policemen or three) as the independent variable. The dependent variable is whether the child hid the boy where no policeman could see him.
That is our guidance, not the authors’.
Participants
The sample is the group of people who take part. The three studies used different children.
- Study One: 10 three-year-olds and 10 four-year-olds.
- Study Two: 30 children aged from 3 years 6 months to 4 years 11 months.
- Study Three: 20 three-year-olds and 20 four-year-olds, in two matched groups (as alike as possible in age and sex), one for each version.
The paper does not say how the children were chosen, so no sampling method can be named. It gives no social class or ethnicity.
Study One: One Policeman
The child sat at a low table. The experimenter said that the policeman was looking for the boy and that the boy wanted to hide.
At first the policeman and the boy faced each other near the edge of the table. The child was asked to “put the wall so that the policeman cannot see the boy”. The wall had to stand on the line of sight.
Then the pair were moved. The line now ran straight out from the table’s edge, at right angles to it. After that, it ran diagonally across the corner.
Piaget and Inhelder (1956) had claimed that young children find a line across a corner especially hard to imagine. In every case the child could see the policeman’s face.
Study Two: Two Policemen
The authors took care that each child understood. So some preliminary questions came first. These were questions to check the child understood the set-up.
The walls were set in the shape of a cross. This made four sections, like four small rooms.
One policeman could see two of the sections but not the other two. The boy was put in each section in turn. The child was asked “can the policeman see the boy?”
The policeman was then moved to the opposite side. The child was asked to hide the boy from him. The exact positions of the policemen are not given here.
If a child made a mistake at these preliminary stages, the error was pointed out and the question repeated until the child was right. Very few mistakes were made, only 8 per cent overall.
A second policeman was then added. “The boy must hide from BOTH policemen.” Now only one section was out of view of both. The child hid the boy. This was repeated three more times, each time leaving a different section as the only hiding place.
Study Three: Harder Versions
The authors made the task harder. The children had found Study Two easy.
In one version, the walls made five sections and there were two policemen. One section was out of view. In the other, there were six sections and three policemen. In this six-section version, the child was told to hide the boy so that “NONE of the policemen can see him”.
Each version had four trials, with the policemen in four different positions. The exact wall layouts are not given here. Each group had 10 three-year-olds and 10 four-year-olds.
A Note on Dates
The study is usually dated 1975, the year of Hughes’ PhD thesis (Hughes, 1975). The details here come from the published report, written with Margaret Donaldson in 1979.
Results
Study One: Almost No Errors
Performance was almost without errors, in the authors’ words.
An egocentric mistake here would be to block the boy from your own view, not the policeman’s.
- Placing the wall: Nine of ten children in each age group placed the wall correctly in every position.
- Across the corner: All the children succeeded when the line ran across the corner of the table.
- Their own view: None placed the wall between themselves and the boy, although he was plainly visible to them.
Being able to see the boy did not seem to affect their judgements.
Study Two: 27 of 30 Children Right
Of the 30 children, 27 (90%) were right on at least three of the four trials.
Twenty-two were right on all four trials. Five more were right on three of the four. The percentages are our arithmetic from the paper’s counts. The paper reports no statistical test for this study.
The younger children did as well as the older ones. The authors concluded that virtually all the children could take account of two points of view at once.
Study Three: Most Children Still Did Well
The table shows how many children made no errors, or one error or none, in each version.
The two rows use different cut-offs. So do not compare the age groups from these counts. Compare each row across.
| Age group | Five sections, two policemen | Six sections, three policemen |
|---|---|---|
| Four-year-olds | 9 of 10 made no errors | 8 of 10 made no errors |
| Three-year-olds | 6 of 10 made one error or none | 7 of 10 made one error or none |
The authors’ own test did compare the age groups. It found the difference significant for the five-section task only. That means it was too big to put down to chance.
The three-policemen task was not harder than the two-policemen task. This surprised the authors, who had predicted that it would be.
What Kinds of Mistakes Did Children Make?
Very few mistakes by the three-year-olds were egocentric. Across the studies, no child consistently hid the boy from their own view instead of the policemen’s.
Two children who failed consistently put the boy in the sections nearest to them. So he was visible to the child and to the policemen.
Two others played a different game. One hid the boy under the table and the other hid him in her hand. The authors scored these as wrong, but only reluctantly. The other mistakes were mostly slips by children who did well otherwise.
Why Was the Hiding Game So Easy?
Human Sense
Why did the children succeed here, when children of this age tended to fail the mountain task? Hughes and Donaldson gave two reasons.
The first is the mountain display itself. It is complicated, because the child has to turn the view round in their mind. Children do better when the display is simpler (Fishbein et al., 1972; Flavell et al., 1968, as cited by Hughes & Donaldson, 1979).
The second is that the child may not understand what to do. The authors say the policemen tasks were extremely clear.
They make human sense: a child understands hiding and seeking, and can imagine being a boy who is hiding. The mountain task has no such point.
Human sense is the authors’ account of this reason. The hiding studies did not test it. A separate 1978 experiment gives some indirect support (Hughes, 1978), described next.
The same research group used this idea in the naughty teddy study, which tested another part of Piaget’s theory (McGarrigle & Donaldson, 1974). It is described on the conservation page.
The Supporting Experiment (Hughes, 1978)
Hughes (1978) tested whether failure on a mountain-style task reflects difficulty understanding what is asked. Three coloured dolls stood on a triangular base, so the child saw one doll’s face and the experimenter saw another. Four-year-olds had to pick the picture showing the experimenter’s view.
Virtually all the children failed the standard version. But 13 of 20 succeeded when the question came after preliminary questions, such as “which doll’s face do I see?” (Hughes, 1978).
Hughes concluded that some failure on a picture task reflects the task and not a missing ability (Hughes, 1978).
The result strongly supports the claim that how a task is presented matters. But the experiment changed the questions, not the meaning. So it does not test human sense directly.
Conclusion
Hughes and Donaldson drew these conclusions.
- Young Children Can Take Another’s View: Three- and four-year-olds worked out what a policeman could see. They ignored what they could see themselves, and kept two or three policemen’s views in mind at once.
- Performance Was High: Performance was “remarkably high”. The authors said the three-year-olds made a noticeable number of mistakes only on the most complex task.
- Piaget Overstated Egocentrism: Other researchers had also found that young children can sometimes work out what another person can see. The authors said this study adds to that evidence, so Piaget overstated egocentrism.
- The Mountain Task May Have Been Hard for Other Reasons: The authors said it was probably hard because the display is complicated and a child may not understand what to do.
But the Study Three table shows the three-year-olds doing slightly better with three policemen than with two (7 of 10 against 6 of 10). So the table does not show three policemen as harder, and it is not clear which task the authors meant.
The study shows that three- and four-year-olds can work out whether someone else can see something. That is not the same skill as the mountain task.
Strengths
The study has six main strengths:
- Tests a Clear Claim: Piaget said young children cannot see another person’s view. On that logic, if they succeed at a hard enough task, his claim goes too far.
- Reliable Results: The result recurred in three studies with different children and with one, two or three policemen. A result that can be repeated is called reliable.
- A Similar Task Gave Similar Results: Young children hid an object from a toy dog. Almost all hid it behind a screen. Moving the screen was harder (Flavell et al., 1978).
- Shows a Limit: When Study Three made the task harder, performance fell, mainly for the three-year-olds and only partly. So the design shows where children start to struggle.
- Clear Measure: The child either hid the boy or did not. Two children who played a different game were scored wrong, but reluctantly.
- Child-Friendly Task: The game had an obvious point, and children often began to respond before the instructions were finished.
We know item 3, a study by Flavell et al., only as reported by Hughes and Donaldson.
Limitations
Before working through each one, here is a one-line summary of the main limitations:
- Different Skill: Not the same job as the mountain task.
- Human Sense Untested: The main explanation was not tested.
- Small Samples: Few children per group.
- Teaching in the Preliminaries: Study Two’s preliminary mistakes were corrected before the tests.
- Artificial Setting: Dolls on a table.
- Generalisability: A small group from the 1970s.
A Different Skill From the Mountain Task
Flavell et al. (1981) distinguished two levels of taking another person’s view. Level 1 asks whether another person can see something. Level 2 asks how it looks from where they are.
In our reading, the policeman tasks fit Level 1 and the mountain task fits Level 2. These are different jobs.
The studies did not give the mountain task to the same children.
So they cannot show that these children could pass it. In Hughes’ own experiment of 1978, virtually all four-year-olds failed a standard mountain-like task (Hughes, 1978).
Human Sense Is Only an Interpretation
No group of children did the same problem without the hiding story.
So we do not know whether the story made the difference. The 1978 experiment gives only indirect support, because it changed the questions asked, not the meaning (Hughes, 1978).
Other explanations stay open.
The mountain display is complicated, so the hiding task may simply be easier. Or the child may check as they go. Some children picked the boy up, moved him and looked round to see whether a policeman could see him.
The authors say they do not know which mental processes were involved. The thesis, read here from a poor scan, leaves open that children used a simple rule, such as avoiding the sections beside a policeman (Hughes, 1975).
Small Samples and Little Statistical Testing
There were 10 children in each group in Studies One and Three, and 30 in Study Two. Study Two had no comparison group.
The paper gives a test result for one comparison only, the age difference in the five-section task, and describes the other as not significant.
Success rates were high.
That may have caused a ceiling effect: a task that is too easy cannot show differences between groups. It may be why three policemen did not look harder than two. That explanation is ours.
Teaching During the Preliminaries
Before the test trials in Study Two, errors were pointed out and questions repeated until the child was right. That is a form of teaching. Hughes (1978) shows that preliminary questions can turn failure into success for most children.
The authors said that mistakes were few, only 8 per cent, and that children often responded before any instruction. Both points weaken the objection. But the paper does not say how the children who made mistakes then did in the test trials.
Artificial Setting and Ecological Validity
Ecological validity means how well findings apply to real life. The children played with toy dolls and walls on a table, which is an artificial set-up. This is our point, not the authors’.
The authors argued that the game made human sense to a child. That is their interpretation. So the study may not show how children cope with real hiding, or with real people’s views.
The children hid a doll in a game run by an experimenter.
Generalisability: Sample and Era
Generalisability means how far the results apply beyond the people, place and time studied. The children were a small group. The paper does not show that they represent all children.
A sample that does not represent everyone shows sample bias. So the ages here are ages for these children, not for all children.
The authors worked at the University of Edinburgh in the 1970s, but the paper gives no recruitment details. So we cannot say where the children were tested. No study described here tested children from other cultures.
Ethical Issues
The paper does not mention consent, parental permission or debriefing. The tasks were games, so harm was probably small.
- Informed Consent: Informed consent means agreeing to take part after being told what will happen. Modern standards would expect a parent’s permission for children this young.
- Right to Stop: Modern standards would also expect the child’s right to stop to be respected.
- Debriefing: Debriefing means explaining the study to participants afterwards. The paper does not describe any.
- Harm: The authors report that the children obviously enjoyed the tasks.
Exam Questions
What a Good Answer Could Include (Our Suggestion From the Specification)
These questions and mark schemes are our own, written in the style of the exam. We do not have an AQA mark scheme for this study. The “Imagine” questions use invented data.
AQA GCSE Psychology (8182) names this study in Paper 1, section 3.1.3, Development. It sits under Piaget’s theory, next to “reduction of egocentricity” and the development of conservation. The other study named there is McGarrigle and Donaldson’s “naughty teddy study”.
AQA expects students to link a study to its theory. It tests knowledge (AO1), application (AO2) and analysis and evaluation (AO3).
Our suggestions:
- Link to Theory: Say what Piaget claimed, what Hughes did instead, and what the result suggests.
- Describe the Study: Name the policemen, the boy doll and the walls, and give a result, such as 27 of 30 children right on at least three of four trials.
- Evaluate It: Use sample size, the artificial setting, whether the children represent others, and ethics. Say what the study shows and what it does not.
Quote the counts in the Results section, such as the Study Three table, not figures from summaries. Many summaries simplify or misapply the paper.
For example, “90 per cent correct” means 27 of 30 children in Study Two were right on at least three of four trials. Only 22 of 30 (73%) were right on all four.
Practice Question 1: Calculate a Percentage (2 marks)
Imagine a study with invented numbers. A researcher asks 20 four-year-olds to hide a doll from one toy policeman. In the first task, 16 children do it correctly. In a harder task, with two policemen, 11 do.
Question: Calculate the percentage of children who were correct in each task. Show your working. [2 marks]
Model answer: first task, 16 ÷ 20 × 100 = 80%. Harder task, 11 ÷ 20 × 100 = 55%.
- Mark scheme: 1 mark for each correct percentage, with working shown.
Practice Question 2: Explain a Difference (2 marks)
Imagine another invented study. Two groups of 20 four-year-olds must work out what a toy policeman can see. Group A does it as a hiding game. Group B does it by choosing a picture. Group A does much better.
Question: Explain why Group A did better. [2 marks]
Model answer: Hughes and Donaldson’s explanation is that a hiding game makes human sense to children. They understand hiding and seeking, so they could work out what the policeman could see. The mountain-style picture task has no such point for a child.
- Mark scheme: 1 mark for human sense, a purpose the child understands. 1 mark for the children then working out what the policeman could see.
Practice Question 3: Evaluate in Two Ways (4 marks)
Question: Explain two ways Hughes’ study can be evaluated. [4 marks]
Model answer: One strength is that the measure was clear. The child either hid the boy or did not, so the scoring rule was simple.
One limitation is the small samples. There were only 10 children in each group in Studies One and Three, so the results may not apply to all children.
- Mark scheme: 2 marks for each way: 1 for naming a point and 1 for explaining it using the study.
- Acceptable Points (Ours): Any two, such as the clear measure, sample size, the artificial setting, ethics or a different skill.
Practice Question 4: Apply to a Scenario (4 marks)
Imagine a nursery teacher shows four-year-olds a model of three hills with a toy bear on the other side. She asks which photograph shows what the bear sees. Most children pick the photograph of their own view. She says that this proves they are egocentric.
Question: Use Hughes’ study to explain why she may be wrong. [4 marks]
Model answer: Hughes and Donaldson used a hiding game, not pictures. Most three- and four-year-olds hid a boy doll where no toy policeman could see him.
On Hughes and Donaldson’s explanation, the teacher’s task may not make human sense to the children. They may not understand what to do. In a hiding game, the same children might show that they can see another’s view.
- Mark scheme: 2 marks for knowledge of Hughes’ study (AO1), such as the hiding game, the result or human sense. 2 marks for applying it to the teacher’s task (AO2).
Practice Question 5: Design an Experiment (6 marks)
Question: Design an experiment to test whether three policemen are harder to hide from than one. [6 marks]
A standardised procedure is one that is the same for every child.
Model answer: I would use three-year-olds. The independent variable is the number of policemen. One group hides the doll from one policeman. The other group hides it from three.
The dependent variable is whether the child hides the doll where no policeman can see it, counted over four trials.
Every child gets the same table, walls and words, with the policemen in the same positions. This is a standardised procedure. It makes the test fair, because only the number of policemen differs.
I expect more errors with three policemen. The child has more to keep in mind.
In Hughes’ Study Three, three policemen were no harder than two. So this expected result is only a prediction for a new design.
- Mark scheme: 1 mark each for the independent variable, the dependent variable, a standardised procedure and why it helps. 2 marks for the expected result (1 prediction, 1 reason).
Practice Question 6: Describe and Evaluate Hughes’ Study (9 marks)
Question: Describe and evaluate Hughes’ policeman doll study. [9 marks]
Model answer (about 270 words):
Hughes and Donaldson tested Piaget’s claim that young children are egocentric. In Piaget’s three mountains task, young children tended to pick their own view of a model, not a doll’s view from another position.
Hughes used a hiding game instead. In Study Two, 30 children aged from 3 years 6 months to 4 years 11 months hid a boy doll from two toy policemen.
Only one section was out of view of both. Of the 30 children, 27 were right on at least three of four trials.
Study Three was harder, with five or six sections and two or three policemen. The three-year-olds made more errors. Most made one error or none. Across the studies, no child consistently hid the boy from their own view.
Hughes and Donaldson concluded that young children can work out what another person can see. They suggested this is because the game made human sense to them.
One strength is that the measure was clear. The child either hid the boy or did not. The result also appeared in three studies with different children.
One weakness is the small samples. There were only 10 children in each group in Studies One and Three. Study Two had no comparison group.
A second weakness is that the study tests a different skill. The policemen tasks ask whether someone can see something. The mountain task asks how a scene looks from another position. The children did not do the mountain task. So the study cannot show that they could pass it.
A third weakness is ethics. The paper does not mention a parent’s permission or debriefing.
Our mark scheme: 4 marks are for describing the study and 5 for evaluating it. The bands below are our own guide, in the style of longer exam questions.
- Describing, 4 marks (AO1): Accurate method, results and conclusion, including the policemen, the boy doll, the walls and a result.
- Evaluating, 5 marks (AO3): Explained points linked to the study, such as sample size, a different skill, ethics or the artificial setting.
- Top band, 7 to 9 marks: Accurate, detailed description and at least two explained evaluation points.
- Middle band, 4 to 6 marks: Mostly accurate description and one or two partly explained evaluation points.
- Bottom band, 1 to 3 marks: Little accurate description, and evaluation that is only named.
Practice Question 7: Ethical Issues and How to Deal with Them (4 marks)
Question: Name two ethical issues in a study like Hughes’. Give one way of dealing with each. [4 marks]
Model answer: One issue is informed consent. The paper does not mention a parent’s permission. A way of dealing with this is to ask a parent for permission first.
Another issue is the child’s right to stop. Modern standards expect it. A way of dealing with this is to tell each child that they can stop the game.
- Mark scheme: 2 marks for each issue: 1 for naming an issue and 1 for a way of dealing with it.
- Acceptable Points (Ours): Issues: informed consent and a parent’s permission, the right to stop, debriefing. Ways: ask a parent, tell the child they can stop, explain the study afterwards.
Later Research
Frick and Baumeler (2017) tested 140 six-year-olds. Those better at stopping an automatic response, called inhibitory control, were better at seeing another person’s view.
This is a correlation, so it does not show cause and effect, and we have read only its summary.
References
Fishbein, H. D., Lewis, S., & Keiffer, K. (1972). Children’s understanding of spatial relations: Coordination of perspectives. Developmental Psychology, 7(1), 21–33. https://doi.org/10.1037/h0032858
Flavell, J. H., Botkin, P. T., Fry, C. L., Wright, J. W., & Jarvis, P. E. (1968). The development of role-taking and communication skills in children. Wiley.
Flavell, J. H., Everett, B. A., Croft, K., & Flavell, E. R. (1981). Young children’s knowledge about visual perception: Further evidence for the Level 1–Level 2 distinction. Developmental Psychology, 17(1), 99–103. https://doi.org/10.1037/0012-1649.17.1.99
Flavell, J. H., Shipstead, S. G., & Croft, K. (1978). Young children’s knowledge about visual perception: Hiding objects from others. Child Development, 49(4), 1208–1211. https://doi.org/10.2307/1128761
Frick, A., & Baumeler, D. (2017). The relation between spatial perspective taking and inhibitory control in 6-year-old children. Psychological Research, 81(4), 730–739. https://doi.org/10.1007/s00426-016-0785-y
Hughes, M. (1975). Egocentrism in preschool children [Doctoral dissertation, University of Edinburgh]. Edinburgh Research Archive. http://hdl.handle.net/1842/22329
Hughes, M. (1978). Selecting pictures of another person’s view. British Journal of Educational Psychology, 48(2), 210–219. https://doi.org/10.1111/j.2044-8279.1978.tb02387.x
Hughes, M., & Donaldson, M. (1979). The use of hiding games for studying the coordination of viewpoints. Educational Review, 31(2), 133–140. https://doi.org/10.1080/0013191790310207
McGarrigle, J., & Donaldson, M. (1974). Conservation accidents. Cognition, 3(4), 341–350. https://doi.org/10.1016/0010-0277(74)90003-1
Piaget, J., & Inhelder, B. (1956). The child’s conception of space. Routledge & Kegan Paul.