Elizabeth Bonawitz

Rutgers, The State University of New Jersey | Rutgers · Department of Psychology (Newark)

Models of pedagogy highlight the reciprocal reasoning underlying learner-teacher interactions, including that learners’ inferences should be shaped by what they believe a teacher knows about them. Yet, little is known about how this influences learning, despite the fact that even young children make rapid inferences about teaching from sparse data....

Teaching is a powerful way to transmit knowledge, but with this power comes a hazard: When teachers fail to select the best set of evidence for the learner, learners can be misled to draw inaccurate inferences. Evaluating others' failures as teachers, however, is a nontrivial problem; people may fail to be informative for different reasons, and not...

Exploratory play supports children’s learning, but the factors that influence play are not fully identified. Here, we investigate whether experiencing an unexpected success on an initial task influences children’s exploration on a subsequent task. In Experiment 1 (N=72), we found that when 4-year-olds successfully completed a puzzle that they were...

We propose that developmental cognitive science should invest in an online CRADLE, a Collaboration for Reproducible and Distributed Large-Scale Experiments that crowdsources data from families participating on the internet. Here, we discuss how the field can work together to further expand and unify current prototypes for the benefit of researchers...

Exploratory play supports children’s learning, but the factors that influence play are not fully identified. Here, we investigate whether experiencing an unexpected success on an initial task influences children’s exploration on a subsequent task. In Experiment 1 (N = 72), we found that when 4-year-olds successfully completed a puzzle that they wer...

Lieder and Griffiths present the computational framework “resource-rational analysis” to address the reverse-engineering problem in cognition. Here we discuss how developmental psychology affords a unique and critical opportunity to employ this framework, but which is overlooked in this piece. We describe how developmental change provides an avenue...

In studies involving human subjects, voluntary participation may lead to sampling bias, thus limiting the generalizability of findings. This effect may be especially pronounced in developmental studies, where parents serve as both the primary environmental input and decision maker of whether their child participates in a study. We present a novel e...

We assessed whether an artifact’s design can facilitate recognition of abstract causal rules. In Experiment 1, 152 three-year-olds were presented with evidence consistent with a relational rule (i.e., pairs of same or different blocks activated a machine) using two differently designed machines. In the standard-design condition, blocks were placed...

Burgeoning evidence suggests that when children observe data, they use knowledge of the demonstrator's intent to augment learning. We propose that the effects of social learning may go beyond cases where children observe data, to cases where they receive no new information at all. We present a model of how simply asking a question a second time may...

Despite limited memory capacity, children are exceptional learners. How might children engage in meaningful learning despite limited memory systems? Past research suggests that adults integrate category knowledge and noisy episodic traces to aid recall when episodic memory is noisy or incomplete (e.g. Hemmer & Steyvers, 2009a,b). We suspect childre...

Burgeoning evidence suggests that when children observe data, they use knowledge of the demonstrator’s intent to augment learning. We propose that the effects of social learning may go beyond cases where children observe data, to cases where they receive no new information at all. We present a model of how simply asking a question a second time may...

We assess whether an artifact’s design can facilitate recognition of abstract causal rules. In Experiment 1, 152 three-year-olds were presented with evidence consistent with a relational rule (i.e., pairs of same or different blocks activate a machine) using a machine with one of two designs. In the standard-design condition, pairs were placed on t...

Constructing an intuitive theory from data confronts learners with a “chicken-and-egg” problem: the laws can only be expressed in terms of the theory’s core concepts, but these concepts are only meaningful in terms of the role they play in the theory’s laws; how can a learner discover appropriate concepts and laws simultaneously, knowing neither to...

Constructing an intuitive theory from data confronts learners with a “chicken‐and‐egg” problem: The laws can only be expressed in terms of the theory's core concepts, but these concepts are only meaningful in terms of the role they play in the theory's laws; how can a learner discover appropriate concepts and laws simultaneously, knowing neither to...

Children are motivated to explore and learn about the world, but they vary in their degree of perseverance during exploration. A growing body of literature suggests that this is malleable from an early age. Here, we ask whether pedagogical questions empower children to persevere during a difficult problem-solving task with a blicket detector machin...

Previous research evaluating the influence of category knowledge on memory found that children, like adults, rely on category information to facilitate recall (Duffy, Huttenlocher, & Crawford, 2006). A model that combines category and target information (Integrative) provides a superior fit to preschoolers recall data compared to a category only (P...

Natural pedagogy emerges early in development, but good teaching requires tailoring evidence to learners’ knowledge. How does the ability to reason about others’ minds support early pedagogical evidence selection abilities? In 3 experiments ( N = 205), we investigated preschool-aged children’s ability to consider others’ knowledge when selecting ev...

Children with more “grit” are more likely to pursue goals despite setbacks; a growing body of research suggests that this “trait” can be learned from an early age. Here, we ask whether certain instructions or questions influence whether a child will attempt to problem solve on their own, or reach out for help. Previous research has shown that when...

Natural pedagogy emerges early in development, but good teaching requires tailoring evidence to learners’ knowledge. How does the ability to reason about others’ minds support early pedagogical evidence selection abilities? In three experiments (N = 205), we investigated preschool-aged children’s ability to consider others’ knowledge when selecting...

For infants and young children, learning takes place all the time and everywhere. How children learn best both in and out of school has been a long-standing topic of debate in education, cognitive development, and cognitive science. Recently, guided play has been proposed as an integrative approach for thinking about learning as a child-led, adult-...

How can education optimize transmission of knowledge while also fostering further learning? Focusing on children at the cusp of formal schooling (N = 180, age = 4.0–6.0 y), we investigate learning after direct instruction by a knowledgeable teacher, after questioning by a knowledgeable teacher, and after questioning by a naïve informant. Consistent...

Question asking is a prevalent aspect of children’s speech, providing a means by which young learners can rapidly gain information about the world. Although past work demonstrates that children are sensitive to the knowledge state of potential informants (e.g., Koenig & Harris, 2005), less work has explored whether children spontaneously direct que...

How do learners’ expectations about teachers’ informativeness shape subsequent learning? Here, we suggest that expectations about teaching style may constrain learning through inferences over (1) the amount of information to be learned, and (2) the importance of the demonstrated information. Adult behavioral data from two experiments conform with o...

This chapter in M. Waldmann (Ed.), Oxford Handbook of Causal Reasoning. Oxford, UK: Oxford University Press, explores the development of causal reasoning in early childhood. We review research on the development of causal reasoning in infancy, toddlerhood, and the preschool years with the broad goals of (1) understanding the origin of our mature ca...

Adults’ causal representations integrate information about predictive relations and the possibility of effective intervention; if one event reliably predicts another, adults can represent the possibility that acting to bring about the first event might generate the second. Here we show that although toddlers (mean age: 24 months) readily learn pred...

Causal learning requires integrating constraints provided by domain-specific theorieswith domain-general statistical learning. In order to investigate the interaction between these factors, preschoolers were presented with stories pitting their existing theories against statistical evidence. Each child heard two stories in which two candidate cause...

Choosing to take certain actions has direct consequences for learning. Do childrentend to choose actions that support learning? We present research suggesting that before reaching kindergarten, children demonstrate proficiency in active learning that cannot be accounted for by simple heuristics for decision-making. Instead, children’s choices revea...

Affective states, exploration, and learning are tightly inter- twined. For example, research has connected surprise to play and learning in early development (Stahl & Feigenson, 2015), but less is known about the potential impact of other affec- tive states and how they might influence exploration and sub- sequent discovery. Given that past researc...

Toddlers readily learn predictive relations between events (e.g., that event A predicts event B). However, they intervene on A to try to cause B only in a few contexts: When a dispositional agent initiates the event or when the event is described with causal language. The current studies look at whether toddlers’ failures are due merely to the diff...

A growing literature suggests that generating and evaluating explanations is a key mechanism for learning and inference, but little is known about how children generate and select competing explanations. This study investigates whether young children prefer explanations that are simple, where simplicity is quantified as the number of causes invoked...

Previous research suggests that three-year-olds fail to learn from statistical data when their prior beliefs conflict with evidence. Are children’s beliefs entrenched in their folk theories, or can preschoolers rationally update their beliefs? Motivated by a Bayesian account, we conducted a training study to investigate this question. Children (45...

Researchers, educators, and parents have long believed that children learn cause and effect relationships through exploratory play. However, previous research suggests that children are poor at designing informative experiments; children fail to control relevant variables and tend to alter multiple variables simultaneously. Thus, little is known ab...

Effective communication entails the strategic presentation of information; good communicators present representative information to their listeners—information that is both consistent with the concept being communicated and also unlikely to support another concept a listener might consider. The present study examined whether preschool-age children...

Different intuitive theories constrain and guide inferences in different contexts. Formalizing simple intuitive theories as probabilistic processes operating over structured representations, we present a new computational model of category-based induction about causally transmitted properties. A first experiment demonstrates undergraduates’ context...

Motivated by computational analyses, we look at how teaching affects exploration and discovery. In Experiment 1, we investigated children’s exploratory play after an adult pedagogically demonstrated a function of a toy, after an interrupted pedagogical demonstration, after a naïve adult demonstrated the function, and at baseline. Preschoolers in th...

People can behave in a way that is consistent with Bayesian models of cognition, despite the fact that performing exact Bayesian inference is computationally challenging. What algorithms could people be using to make this possible? We show that a simple sequential algorithm “Win-Stay, Lose-Sample”, inspired by the Win-Stay, Lose-Shift (WSLS) princi...

Young children often endorse explanations of the natural world that appeal to functions or purpose—for example, that rocks are pointy so animals can scratch on them. By contrast, most Western-educated adults reject such explanations. What accounts for this change? We investigated 4- to 5-year-old children’s ability to generalize the form of an expl...

We present a proposal—“The Sampling Hypothesis”—suggesting that the variability in young children’s responses may be part of a rational strategy for inductive inference. In particular, we argue that young learners may be randomly sampling from the set of possible hypotheses that explain the observed data, producing different hypotheses with frequen...

Although probabilistic models of cognitive development have become increasingly prevalent, one challenge is to account for how children might cope with a potentially vast number of possible hypotheses. We propose that children might address this problem by ‘sampling’ hypotheses from a probability distribution. We discuss empirical results demonstra...

Research suggests that the process of explaining influences causal reasoning by prompting learners to favor hypotheses that offer “good” explanations. One feature of a good explanation is its simplicity. Here we investigate whether prompting children to generate explanations for observed effects increases the extent to which they favor causal hypot...

Bayesian models have been applied to many areas of cognitive science including vision, language, and motor learning. We discuss the implications of this framework for cognitive development. We first present a brief introduction to the Bayesian framework. Bayesian models make assumptions about representation explicit, and provide a detailed account...

In the article we argue that past Bayesian approaches that model children's learning from data are missing an important element — the role of other people in generating that data. We propose that children take the origin of data into account when learning, which can be understood through ideal observer analyses of the social situation. Moreover, wh...

How can education optimize transmission of knowledge while also fostering further learning? Focusing on children at the cusp of formal schooling (N = 180, age = 4.0 - 6.0 y), we investigate learning after direct instruction by a knowledgeable teacher, after questioning by a knowledgeable teacher, and after questioning by a naïve informant. Consiste...

Questioning is a core component of formal pedagogy. Parents commonly question children, but do they use questions to teach? Here we define “pedagogical questions” as questions for which the questioner already knows the answer and intended to help the questionee learn. We investigate the frequency and distribution of pedagogical questions from paren...

Choosing to take certain actions has direct consequences for learning. Do children tend to choose actions that support learning? We present research suggesting that before reaching kindergarten, children demonstrate proficiency in active learning that cannot be accounted for by simple heuristics for decision-making. Instead, children’s choices reve...

We look at the effect of evidence and prior beliefs on exploration, explanation and learning. In Experiment 1, we tested children both with and without differential prior beliefs about balance relationships (Center Theorists, mean: 82 months; Mass Theorists, mean: 89 months; No Theory children, mean: 62 months). Center and Mass Theory children who...

Young children often endorse explanations of the natural world that appeal to functions or purpose, explaining (for example) that rocks are pointy so animals can scratch on them. By contrast, most Western-educated adults reject such explanations. What accounts for this change? We investigated 4- to 5-year-old children’s ability to generalize the fo...

Questioning is a core component of formal pedagogy. Parents commonly question children, but do they use questions to teach? This article defines "pedagogical questions" as questions for which the questioner already knows the answer and intended to help the questionee learn. Transcripts of parent-child conversations were collected from the CHILDES d...

Research suggests that the process of explaining influences causal reasoning by prompting learners to favor hypotheses that offer “good” explanations. One feature of a good explanation is its simplicity. Here, we investigate whether prompting children to generate explanations for observed effects increases the extent to which they favor causal hypo...

In the article we argue that past Bayesian approaches that model children's learning from data are missing an important element — the role of other people in generating that data. We propose that children take the origin of data into account when learning, which can be understood through ideal observer analyses of the social situation. Moreover, wh...

Effective communication entails the strategic presentation of information; good communicators present representative information to their listeners-information that is both consistent with the concept being communicated and also unlikely to support another concept a listener might consider. The present study examined whether preschool-age children...

Bayesian models have been applied to many areas of cognitive science including vision, language, and motor learning. We discuss the implications of this framework for cognitive development. We first present a brief introduction to the Bayesian framework. Bayesian models make assumptions about representation explicit, and provide a detailed account...

People can behave in a way that is consistent with Bayesian models of cognition, despite the fact that performing exact Bayesian inference is computationally challenging. What algorithms could people be using to make this possible? We show that a simple sequential algorithm "Win-Stay, Lose-Sample", inspired by the Win-Stay, Lose-Shift (WSLS) princi...

Although probabilistic models of cognitive development have become increasingly prevalent, one challenge is to account for how children might cope with a potentially vast number of possible hypotheses. We propose that children might address this problem by 'sampling' hypotheses from a probability distribution. We discuss empirical results demonstra...

Probabilistic models of cognitive development indicate the ideal solutions to computational problems that children face as they try to make sense of their environment. Under this approach, children's beliefs change as the result of a single process: observing new data and drawing the appropriate conclusions from those data via Bayesian inference. H...

Probabilistic models of cognitive development indicate the ideal solutions to computational problems that children face as they try to make sense of their environment. Under this approach, children's beliefs change as the result of a single process: observing new data and drawing the appropriate conclusions from those data via Bayesian inference. H...

We present a proposal-"The Sampling Hypothesis"-suggesting that the variability in young children's responses may be part of a rational strategy for inductive inference. In particular, we argue that young learners may be randomly sampling from the set of possible hypotheses that explain the observed data, producing different hypotheses with frequen...

Abstract—An intuitive theory is a system of abstract concepts and laws relating those concepts that together provide a framework for explaining some domain of phenomena. Constructing an intuitive theory based on observing the world, as in building a scientific theory from data, confronts learners with a “chickenand-egg” problem: the laws can only be...

We look at the effect of evidence and prior beliefs on exploration, explanation and learning. In Experiment 1, we tested children both with and without differential prior beliefs about balance relationships (Center Theorists, mean: 82 months; Mass Theorists, mean: 89 months; No Theory children, mean: 62 months). Center and Mass Theory children who...

Toddlers readily learn predictive relations between events (e.g., that event A predicts event B). However, they intervene on A to try to cause B only in a few contexts: When a dispositional agent initiates the event or when the event is described with causal language. The current studies look at whether toddlers' failures are due merely to the diff...

Previous research suggests that 3-year-olds fail to learn from statistical data when their prior beliefs conflict with evidence. Are children's beliefs entrenched in their folk theories, or can preschoolers rationally update their beliefs? Motivated by a Bayesian account, we conducted a training study to investigate this question. Children (45 mont...

A growing literature suggests that generating and evaluating explanations is a key mechanism for learning and inference, but little is known about how children generate and select competing explanations. This study investigates whether young children prefer explanations that are simple, where simplicity is quantified as the number of causes invoked...

Motivated by computational analyses, we look at how teaching affects exploration and discovery. In Experiment 1, we investigated children's exploratory play after an adult pedagogically demonstrated a function of a toy, after an interrupted pedagogical demonstration, after a naïve adult demonstrated the function, and at baseline. Preschoolers in th...

How does explicit instruction affect exploratory play and learning? We present a model that captures pedagogical assumptions (adapted from Shafto and Goodman, 2008) and test the model with a novel experiment looking at 4-year-olds' exploratory play in pedagogical and non-pedagogical contexts. Our findings are consistent with the model predictions:...

Bayesian models of cognition are typically used to describe human learning and inference at the computational level, iden-tifying which hypotheses people should select to explain ob-served data given a particular set of inductive biases. However, such an analysis can be consistent with human behavior even if people are not actually carrying out exa...

Adults' causal representations integrate information about predictive relations and the possibility of effective intervention; if one event reliably predicts another, adults can represent the possibility that acting to bring about the first event might generate the second. Here we show that although toddlers (mean age: 24 months) readily learn pred...

In this thesis, rational Bayesian models and the Theory-theory are bridged to explore ways in which children can be described as Bayesian scientists. I investigate what it means for children to take a rational approach to processes that support learning. In particular, I present empirical studies that show children making rational predictions, expl...

Researchers in both educational and developmental psychol-ogy have suggested that children are not particularly adept hy-pothesis testers, and that their behavior can often appear irra-tional. However, a growing body of research also suggests that people do engage in rational inference on a variety of tasks. Recently researchers have begun testing...

Some researchers have suggested that correlation information and information about action are bound in a single representation: "causal knowledge". If children have only observed correlation information, do they spontaneously try to generate the effect? Do they represent the relationship as potentially causal? We present three action and looking-ti...

Previous research (e.g., S. A. Gelman & E. M. Markman, 1986; A. Gopnik & D. M. Sobel, 2000) suggests that children can use category labels to make inductive inferences about nonobvious causal properties of objects. However, such inductive generalizations can fail to predict objects' causal properties when (a) the property being projected varies wit...

Different intuitive theories constrain and guide inferences in different contexts. Formalizing simple intuitive theories as probabilistic processes operating over structured representations, we present a new computational model of category-based induction about causally transmitted properties. A first experiment demonstrates undergraduates' context...

A growing literature suggests that generating and evaluating explanations is a key mechanism for learning and development, but little is known about how children evaluate explanations, especially in the absence of probability information or robust prior beliefs. Previous findings demonstrate that adults balance several explanatory virtues in evalua...

In this paper we show that, given identical evidence, children with different naïve theories exhibit different patterns of exploratory play. Karmiloff-Smith & Inhelder (1974) demonstrated that before children develop an adult "Mass Theory" of balance, they entertain a "Center Theory", believing that all objects should be balanced at their geometric...

Human intelligence has long inspired new benchmarks for research in artificial intelligence. However, recently, research in machine learning and AI has influenced research on children's learning. In particular, Bayesian frameworks capture hallmarks of children's causal reasoning: given causally ambiguous evidence, prior beliefs and data interact. H...

Causal learning requires integrating constraints provided by domain-specific theories with domain-general statistical learning. In order to investigate the interaction between these factors, the authors presented preschoolers with stories pitting their existing theories against statistical evidence. Each child heard 2 stories in which 2 candidate c...

Researchers, educators, and parents have long believed that children learn cause and effect relationships through exploratory play. However, previous research suggests that children are poor at designing informative experiments; children fail to control relevant variables and tend to alter multiple variables simultaneously. Thus, little is known ab...

Very young children have remarkably sophisticated causal knowledge about the world, yet relatively little is known about the process of causal learning. In this paper we provide a Bayesian model of how the interaction of prior theories and evidence can lead to ambiguity in competing causal hypotheses; we suggest that children seek to resolve such a...

We propose a rational analysis of children's false belief reasoning. Our analysis realizes a continuous, evidence- driven transition between two causal Bayesian models of false belief. Both models support prediction and ex- planation; however, one model is less complex while the other has greater explanatory resources. Because of this explanatory a...

Although there is widespread agreement that children learn through exploratory play, little is known about the factors that affect children's exploratory play or how exploratory play might lead to accurate learning. However, recent research suggests that children's exploratory play might be affected by the ambiguity of the evidence children observe...

Previous work has demonstrated the importance of both naïve theories and statistical evidence to children's causal reasoning. In particular, four-year-olds can use statistical evidence to update their beliefs. However, the story is more complex for three-year-olds. Although three-and-a-half-year-olds perform as well as four-year-olds when statistic...

Different kinds of knowledge are relevant in different inductive contexts. Previous models of category-based induction have focused on judgments about taxonomic properties, but other kinds of models are needed for other kinds of properties. We present a new model of reasoning about causally transmitted properties. Our first experiment shows that th...

This study investigates the interaction between preschoolers' initial theories and their ability to learn causal relations from patterns of data. Children observed ambiguous evidence in which sets of two candidate causes co-occurred with an effect (e.g. A&B E, A&C E, A&D E, etc). In one condition, all candidate causes were from the appropriate doma...