MY RESEARCH AND WRITING

Brown University Pre-Baccalaureate Course / Summer Semester 2025 / Prof. Sobel

Did I Just Learn This Myself?

Examining Children’s Ability to Understand Their Own Patterns of Learning

Summary 

In their 2024 research study Two-year-olds can reason about the temporal structure of their performance, Peter Zhu, Veronica Aranda, Grace Keene, and Hyowon Gweon found that children as young as two years old can understand their own act of learning through the particular sequence of failure and success in given tasks. Their experiments found that children who had success at tasks after failure were more likely to try again than those who succeeded the same amount but randomly. The study added onto previous studies that demonstrated this type of awareness in older children but not children as young as this experiment observed. By building onto these findings that this learning is present in children younger than previously believed, this paper proposes an experiment that furthers this research by studying the influence of scaffolding or help in succeeding at tasks and how children view their own progress as a result.

Why Study This?

Much of the research on the cognitive development of children focuses on how children think and learn, and when they are able to do so. Important work by some of the leading figures in child developmental psychology, from Jean Piaget’s foundational theories tracking the different stages of cognitive development, Renée Baillargeon’s significant research on infant cognition, Elizabeth Spelke’s ideas about children’s innate knowledge, and Alison Gopnik’s studies on how young children understand others, have all contributed to our understanding of how children learn. However, another important topic is children’s ability to recognize that they are actually learning, or, in other words, learning about learning. As we develop from childhood to adulthood, knowing what one is good at and the progress one makes in specific tasks is very important in helping us decide when and how to persevere in challenging situations or those that we want to pursue. Being able to understand when someone is getting better at something plays an important role in decision-making and self-reflection. Researchers in the field of child development have conducted a lot of work to measure when this type of understanding first develops and how it continues to develop in the first few years of life.

As mentioned by Sobel and Letourneau in their 2015 study, there is still a lot to be understood about how young children become aware that they are learning and processing new information. In studying 4-10 year old children, Sobel and Letourneau found that they had an understanding of their learning through interviewing children and asking them to reflect on their experiences with learning new knowledge.

In 2023, Julia Leonard and her colleagues also studied 4-6 year old children and found that they were more interested in attempting easier tasks if they had repeated successes than significantly harder tasks. The results showed that children were more likely to try more difficult tasks when they saw evidence that their learning improved over time. This study, however, focused on children who were 4 years old and older. It showed that they can track the progress of their performance in certain tasks over time. However, it also raised the questions about if these capabilities were present in even younger children. An earlier study by Leonard (2017) found that 13-18 month old infants were more persistent at tasks when they saw adults having success with them. This suggests that at an early age, there is a foundational understanding of learning about learning being built.

To study if and how children younger than four years of age could understand their own learning patterns, studies have to minimize the role of complex language acquisition, as that is one of the barriers in studying very young children. While some studies track eye movements to determine whether or not children have complex language skills, this type of measurement would not be best suited to understand if they understand that they are learning. Eye tracking is helpful for measuring young children’s attention and preference, but not their own reflection on tasks they complete. Similar to language development, motor skill development is also growing at a very young age. Studies that measure an understanding of children’s own learning need to involve basic skills and tasks that a child is able to perform.

The Research Hypothesis 

Peter Zhu and his colleagues were interested in the idea of whether or not children as young as two-years old could recognize that they were actually learning and improving in tasks, and how it influenced their reactions. Based on prior studies that tested children as young as four-years old, they theorized that two- to three-year old children could also understand and track how they perform and learn. Their main hypothesis centered around the theory that the ability to reason and think about one’s own learning is present at a very early age.

How Did They Test This?

In order to test their hypothesis, the researchers conducted two experiments to better understand how and when children’s temporal reasoning about their own progress emerges. Experiment 1 was a preliminary study, which was used to measure children’s initial response to the testing of their hypothesis in order to be able to make informed decisions about any adjustments they wanted to make for the second experiment. With sample data from Experiment 1 that they felt would reinforce their assumptions, the researchers could make changes to the group or task to refine the testing in Experiment 2.

Experiment 1

In Experiment 1, children were given two colorful toys that had buttons and lights, and the task of pressing the buttons to make it light up. The researchers needed to make sure that the study was not influenced by children’s own physical ability to solve the task, but rather their understanding of their own success. To do this, they made the task easy but controlled the failures and successes of activating the light through a hidden remote control under the table. The researchers kept verbal instructions to a minimum to ensure that language acquisition was not a factor in completing the task. The first toy, which was a Control Toy, allowed children to successfully light up the toy three times in a row. The researchers believed getting children accustomed to the task, and success, was important. Then, children were shown a Test Toy and attempted to light up six times, where researchers controlled the sequence of failures and successes. The children were divided into two groups. The first group failed and succeeded at making the toy light up in a seemingly random order, SSFFSF, controlled by the experimenter. A second group was made to fail three times followed by three successes, FFFSSS, which showed direct improvement over time. Both groups succeeded and failed three times, but the order was manipulated to test if it affected how children viewed their successes. Afterwards, children were asked which toy they wanted to show their parents.

Variables

The independent variable of the study, or what the experimenters changed for the two groups, was controlling the order of a child’s success and failure at the task. All elements of the task were similar except for the order of successes and failures. The researchers theorized that this single variable would influence how children viewed their progress, and isolated it while keeping all other variables the same. The pattern of successes for children was critical for this study, because it could show their understanding that they learned how to succeed at lighting up the toy as opposed to the group with random successes that may feel less confident in having learned how to make it work.

The dependent variable, or what researchers measured, was the choice of toy box the children selected to show their parents. They could choose the Control Toy, which they succeeded three times straight in lighting up or the Test Toy, which they succeeded in lighting up half the time.

Results

Experiment 1 tested 16 two- and three- year old children. Out of the 16 children, 8 were given the random success (SSFFSF), and called the stochastic group, and 8 were allowed to improve (FFFSSS), and called the improvement group. As hypothesized, the researchers found that children in the improvement group had a higher rate (8 of 8, or 100%) of choosing the test box to show their parents. Only half of the stochastic group (4 out of 8, 50%) chose the Test Toyin which they succeeded just as much as the other group, but not in an order that would simulate success in learning how to make it work.

Experiment 2

Experiment 2 utilized the same test but applied changes based on the initial learnings from the smaller test group in experiment 1. The researchers found no major differences in how 2- and 3-year olds reacted, and decided to test only two-year olds in the second experiment. They also increased the sample size to 60 participants. Based on the results of the first experiment, researchers also decided prior to the second experiment to exclude children in the final results, based on various criteria. Following Experiment 1, which was a preliminary study, Experiment 2 was a pre-registered study. To ensure overall accuracy, Experiment 2 excluded participants whose actions could have resulted in possible errors and inconsistencies in the data. This included failing to finish the task, technical or experimentor errors, interference by parents, and taking too long in making a decision. 

The primary hypothesis for Experiments 2 was the same: more children would be more likely to show their parents a toy with which they improved after failure, instead of one that gave them success and failure in an order that did not show them an improvement. Researchers still expected children in the random group to be hesitant in choosing the Test Toy over the Control Toy when their experience with the Test Toy varied more through random successes, was more unpredictable, or was harder to see consistent improvement.

As the procedures and tasks stayed constant in both experiments 1 and 2, the independent variable remained the order of the successes and failures for the children in the improvement condition and the stochastic condition. The dependent variable also remained unchanged in experiment two, as the Control Toy and Test Toy were identical.

The results of Experiment 2 showed a greater preference of the Test Toy by the improvement group over the stochastic group. 20 out of 30 children in the improvement group chose the Test Toy, while only 9 out of 30 from the stochastic group chose the Test Toy. The results still show a clear preference for the Test Toy in the improvement group. It further highlighted the preference of children in the stochastic group for the success they found in the Control Toy, as opposed to the Test Toy that gave them mixed results.  Although the percentages for both groups were lower in experiment two as opposed to experiment 1, the researchers found it to be evidence to support their initial hypothesis. The results of the study confirmed the hypothesis and provided empirical evidence that even 2-year old children were aware of a positive pattern in their learning and applied that knowledge to their decisions.

Proposed Experiment That Builds on This Research

While Peter Zhu and his colleagues’ study highlighted that children as young as two years old can understand and think about how their progress changes over time, the results open up many other interesting and related questions to study. This proposed follow-up experiment to Peter Zhu’s study tests the potential impact of scaffolding, or adults helping children achieve success, and if this influences their own reasoning about success in their task. Do children value their own independent success more or less than when helped by an adult? In this study, researchers will test how actively helping and assisting children in successfully completing their tasks affects children’s awareness of their own performance and learning. The overall purpose of this study is to determine if children base their understanding of progress on their own improvement over time or on external factors, such as assistance from an adult. It mirrors the structure of Peter Zhu’s experiment but replaces the choice between random success and learned success with adult scaffolding to see if receiving help in achieving success changes children’s perception of their success. The potential findings will help researchers understand if adults shape children’s awareness of their learning positively or if they interfere negatively with it.

Participants

This study will test 80 three-year old children with no abnormal developmental, cognitive, or motor skill delay. They will be selected from various pre-schools and/or children’s museum playgroups. It tests children of similar age as Peter Zhu’s study to provide further empirical evidence for this age group and allows comparisons to his data directly. This study focuses on three-year olds as they may have a stronger idea about and sense of their own accomplishment than two- year olds, which is relevant to this study. The entire group of 80 will be separated into two- groups of 40 children. The children will need to be able to understand and respond to prompts, as well as functionally complete a basic skill test. The groups will have to understand verbal commands and have basic motor skills to complete the task, even though children’s successes or failures are controlled remotely by the experimenter. 

Materials and Stimuli

Materials needed for this study are two boxes that have buttons and lights. To build directly on Peter Zhu et al’s study, this test will use the same control and Test Toys they used. The small boxes are built out of foam boards and felt, with colorful buttons and a blinking LED light that is controlled by a hidden remote control by the administrator of the test. To easily separate the boxes for identification, the color of each is different. However, neither box should be more visually interesting or novel than the other. Parents should also be nearby so when the child gets asked which toy they wish to show their parents, it is believable.

Procedure

All children are asked to play with two toy boxes, the Control Toy and the Test Toy. In phase 1, which is the control phase, all children are successful at making the toy light up three times when pressing a button on the box, which is controlled by the experimenter. The children are succeeding on their own without failure, and there is no scaffolding or help given to either of the two groups.

In phase two, children are given the Test Toy to play with. Half of the children, who are in Group A, are asked to make the light work on their own, while the other half, who are in Group B, are given help in doing so. Both groups fail on their first three attempts and succeed in their next three tries. Children in Group A succeed on their own three times after failure three times with no help from the experimenter. Children in Group B fail three times and require help from the experimenter to be able to succeed three times. The help from the experimenter involves suggesting which button to press and when to do so in order to make the light go on. Their help is needed for the child to be able to succeed at the task. At the end of the experiment, the children are asked: “which of the toy boxes do you want to show your parents when they enter the room?”

The independent variable in this study is whether or not the child is given help in succeeding at making the Test Toy light up. The dependent variable is the toy box that is chosen by the children. This is key to determining if the children’s understanding of their own success is impacted by support from someone else. For replication, it is important to maintain consistency in the specific type of help or assistance from the experimenter in the form of saying “This is how you press it. This is how you light it up. Press this button.” It should be direct help in the form of verbal instruction.

Hypothesis

Three-year olds who receive no help during a learning task and improve on their own will be slightly better at identifying their own improved performance over time than those who received help from the experimenter. It would be expected for children who succeeded independently to choose the Control Toy more than those who received help. If the results demonstrate this, it can be inferred that children will be more aware that they improved based on their own abilities, as opposed to the children who received help. If the data of this experiment showed the opposite, and that more children chose the toy with which they received help, it may point to the social interaction and support having a positive influence on learning, or the positive feeling that their learning was made easier. The added involvement of the experimenter in Group B may influence the results as children enjoy socialization and chose the toy that involved more social interaction. Either way, it would be interesting to see the results because scaffolding is an important part in building a young child’s learning potential. However, this is not at the expense of the value of independent learning. Questions that remain range from the timing of scaffolding, the type of help, whether it is physical or verbal, the impact of social interaction has on guided learning, and if these results vary by age. It is an interesting topic that encourages further research.

APA Citations

1. Zhu, P., Aranda, V., Keene, G., & Gweon, H. (2024). Two-year-olds can reason about the temporal structure of their performance. Proceedings of the Annual Meeting of the Cognitive Science Society, Volume 46. https://escholarship.org/uc/item/1vb912rp.
2. Leonard, J., Cordrey, S., Liu, H., & Mackey, A. (2023). Young children calibrate effort based on the trajectory of their performance. Developmental Psychology, Volume 59, 609-619. https://psycnet.apa.org/record/2023-05428-001.
3. Leonard, J., Lee, Y., & Schulz, L. (2017). Infants make more attempts to achieve a goal when they see adults persist. Science, Volume 357, 1290-1294. https://www.jstor.org/stable/26399892
4.  Sobel, D., & Letourneau, S. (2015). Children’s developing understanding of what and how they learn. Journal of Experimental Child Psychology, Volume 132, 221-229. https://www.sciencedirect.com/science/article/pii/S0022096515000168. 

NEXT PROJECT