Testing a Framework of Environmental Adaptation in Children's Learning Strategies

PI/PD：凯瑟琳亚当斯香农 项目摘要 儿童在早期社会环境中是如何学习的？社会学习的一个重要因素是 来自社会伙伴的帮助的响应性或可靠性。出现反应迟钝或 不一致的解释可能经常会让他们的帮助得不到回应;其他孩子可能会经历 不同的社会伙伴在帮助他人方面各不相同。尽管它对长期的学习成果有影响， 我们仍然不太了解幼儿如何调整他们的学习策略，作为可靠性的函数， 帮助他们的社会环境，以及这种适应如何塑造学习成果。 当前提案的目标是开发和测试一个描述环境 儿童学习策略的适应性。在一系列针对4至6岁儿童的实验中，我们从一个 简单的设计将“帮助的可靠性”可操作化，即社交伙伴是否始终如一地做出响应 或者在儿童完成学习任务之前忽略帮助的请求（例如，了解玩具是如何工作的）。在这 实验设置，我们解决3个具体目标： 在目标1中，我们问孩子们是否适应他们的（1）学习目标和（2）决定独自探索或寻求 帮助基于过去的社会合作伙伴的帮助，如果是这样，这些适应是否最大限度地学习 结果是得到帮助的可能性。如果一种行为有利于在一个社会环境中学习， 环境，但同样的行为可能导致另一种成本。因此，有益的学习适应 如果环境发生变化，它们会变得适应不良。在目标2中，我们引入了一个 可靠到不可靠和不可靠到可靠的变化，例如， 在学习停止对孩子的帮助做出反应之前（反之亦然），在这里，我们测试是否 环境的变化会导致孩子使用不匹配的学习策略，这反过来会导致 次优的学习成果。在目标3中，我们构建了一个描述环境的计算模型， 根据基于效用的社会推理原则，调整儿童的学习策略，其中 儿童考虑其社会伙伴的行为以最大化预期效用（即，净收益 成本）。我们在实验中测试模型预测对儿童的行为。 总之，这项工作将提供关键的经验数据和计算模型，以推进我们的研究。 了解儿童如何在不同的社会背景下学习。这项建议的长远目标是 对儿童，特别是那些生活在缺乏支持的环境中的儿童， 在社会学习过程中做出决定，塑造他们未来的学习成果。研究结果 承诺为改善社会学习互动的干预措施提供信息，作为支持早期教育的一种手段。

PI/PD: Katherine Adams Shannon PROJECT SUMMARY How do children learn within their early social environments? An important factor in social learning is the responsiveness or reliability of help from social partners. Children who experience unresponsive or inconsistent caregiving may often have their bids for help go unanswered; other children may experience different social partners who vary in their helpfulness. Despite its ramifications for long-term learning outcomes, we still understand little about how young children adapt their learning strategies as a function of the reliability of help in their social environments, and how such adaptations shape learning outcomes. The goal of the current proposal is to develop and test a framework that describes environmental adaptation in children’s learning strategies. In a series of experiments with 4- to 6-year-olds, we start with a simple design that operationalizes ‘reliability of help’ as whether or not a social partner consistently responds to or ignores bids for help before children complete a learning task (e.g., figure out how a toy works). Within this experimental setting, we address 3 specific aims: In Aim 1, we ask if children adapt their (1) learning goals and (2) decisions to explore alone or seek help based on the past helpfulness of a social partner, and if so, whether these adaptations maximize learning outcomes given the likelihood of receiving help. A given behavior is adaptive if it benefits learning in one social environment, but the same behavior can lead to costs in another. It follows that a beneficial learning adaptation can become maladaptive if the conditions of the environment change. In Aim 2, we introduce a reliable-to-unreliable and unreliable-to-reliable change in the reliability of help, such that a social partner who was helpful before learning stops responding to children’s bids for help (and vice versa). Here, we test whether an environmental change can lead children to use mismatched learning strategies that, in turn, lead to suboptimal learning outcomes. In Aim 3, we construct a computational model that describes environmental adaptation in children’s learning strategies according to the principles of utility-based social reasoning, wherein children consider the behavior of their social partners to maximize the expected utility (i.e., benefits net of costs) of a given learning decision. We test model predictions against children’s behavior in the experiments. Taken together, this work will provide key empirical data and a computational model to advance our understanding of how children learn within diverse social contexts. The long-term objective of this proposal is to generate novel insights into the ways in which children, particularly those who live in less supportive environments, make decisions during social learning that shape their future learning outcomes. Findings have promise to inform interventions that improve social learning interactions as a means to support early education.