Although many studies have shown that being exposed toempirical data that contradict one’s beliefs can lead to learning,it is not clear whether calling attention to inconsistenciesamong beliefs without the provision of new data, leads tolearning. The present study asked whether calling attention toinconsistent beliefs via thought experiments leads to beliefrevision. Five-hundred-seventy-five participants were assignedto three different conditions in a pre-training, training, post-training design. The results showed that participants generatedinconsistent beliefs between pre-training and training, but theydid not spontaneously revise them at post-training (BaselineCondition). They did revise them, however, when they wereasked to reason about the implications of the training thoughtexperiments (Warning Condition) and when they saw anexplicit inference drawn from the training thought experiments(Explicit Inference Condition). These results show that, withprompting, scientifically naïve adults can learn from thoughtexperiments.

As the history of science has documented, there is an important role for thought experiments in scientific progress. Yet, there is very little empirical research about whether and how children learn from thought experiments. Here, we asked that question in the context of 6-year-olds’ developing theory of matter. At the outset of the study, over half of the children claimed that small pieces of matter weigh nothing at all. Children were randomly assigned to a Real (RE) and a Thought Experiment (TE) condition. The goal of each condition was to show – via demonstration in the RE and via mental simulation in the TE – that the weight of a single grain of rice can cause a card resting on a fulcrum to topple. We found that children simulated accurately in the TE, and they changed their judgments and justifications concerning the weight of small pieces equally from the TE and RE.

Questions are prevalent in everyday speech and they are often used to teach. When learners receive explicit cues that the intent of a question is pedagogical, they draw inferences that lead to superior learning. Although the ability to infer pedagogical intent is critical, very little is known about the mechanisms that support the inference that any particular statement is pedagogical or not. We tested the hypothesis that the prosody of speech marks the intent of pedagogical and information-seeking questions. In Studies 1 and 2, 256 naïve participants rated 100 pedagogical and information-seeking questions, spoken in child- or adult-directed speech. We found that naïve listeners can accurately infer pedagogical intent on the basis of prosody alone. In Study 3, we begin charting the acoustic features that differentiate pedagogical from information-seeking questions. These findings provide a window into the mechanisms that allow learners to infer pedagogical intent in otherwise ambiguous situations.

Research has shown that pragmatic, social, and prosodic cues are used to infer the communicative intent of a speaker,including pedagogical intentions (Bohn & Frank, 2020; Cristia, 2013; Csibra & Gergely, 2009). However, little is knownabout whether prosodic features can signal pedagogical intent in syntactically equivalent utterances. We asked whetherprosodic features can carry information about the intent of a question (i.e., whether it is a pedagogical or an informationseeking question), both within child- and adult-directed speech. Eighty naive participants were asked to classify questionsgenerated by five different speakers. We found that participants could reliably discriminate between questions intendedto be pedagogical from those that were intended to be information seeking, both within child- and adult-directed speech,although pedagogical questions were detected more successfully when spoken with child-directed speech. These findingsindicate that prosody may convey pedagogical intent, which in turn may facilitate learning.

Although children are on the receiving end of pedagogical questions (asked with an intent to teach) and information-seeking questions (asked with an intent to seek information), little is known about how children differentiate between the two types of questions. Here, we tested if parents spontaneously modulate their prosody when asking pedagogical as opposed to information- seeking questions. To test this, we asked 35 parent- child pairs to engage in a learning game where parents were asking questions while being in the role of a teacher (pedagogical) or in the role of a student (information- seeking). Next, 128 naïve listeners judged the questions produced by parents. We found that naïve listeners could reliably differentiate the two types of questions on the basis of prosody alone. This finding highlights the importance of prosody as a mechanism for communicating pedagogical intent in parent-child interactions.

Although the ability to recognize pedagogical intent is important, very little is known about how children do it in ambiguous contexts. For example, when the context is unambiguous, children who receive pedagogical questions (where the intent is to teach) show superior learning compared to children who receive information-seeking questions (where the intent is to gather information) (Yue et al., 2018). Here, we asked if children could use the prosody of speech to recognize pedagogical intent. A total of 98 children (ages 4 to 6), received 24 trials where in each trial the same question was asked by two individuals: one with pedagogical and one with information-seeking prosody. We found that both 5- and 6-year-olds could recognize “who the teacher is,” while 4-year-olds performed at chance. Age was a significant predictor of performance. These results suggest that prosody can serve as an important mechanism for recognizing pedagogical intent in early childhood.

Affective components are frequently overlooked in computational modelling, despite the notable role of emotions in learn-ing. Towards the goal of measuring affect in learning, we developed a theory-based Bayesian model that predicts surprisebased on a learners prior beliefs and the evidence observed, and then compared the model to a physiological measure com-monly suggested to capture surprise: pupil dilation. Critically, we also investigate whether this correlation is strong whenparticipants predict the events. Comparing our model predictions to the first four test trial responses from 93 participants(mean age: 8.00 years) revealed a significant, positive correlation when making predictions (r(9)=.55, p=0.04), a negativecorrelation when only evaluating (r(9)=-.50, p=0.07), and significant difference between groups (z=2.34, p¡0.01). Nextsteps will allow us to build on this result by developing a modified Bayesian model, that takes physiological surprise as acomponent in predicting the participants learning.