认知与发展心理学研究室知识库

客观世界包含着丰富且复杂的因果现象。在成长过程中，儿童面对的因果问题也越来越复杂。儿童依赖两种方式学习因果知识——直接经验和他人的陈词。为了学习新颖的、没有直接证据的因果关系，儿童需要依赖他人获取知识。学前儿童通过他人获取知识的过程是选择性的，他们并非被动地全盘接受他人提供的信息，而是有选择地采信。客观世界不仅提供了可直接观察因果关系的物理证据，还提供了间接学习的社会参照线索。以往研究发现，学前儿童可以根据这些线索评估信息源的可信度，进而选择向可靠的信息源学习新知识。

信息充分性是儿童可利用的线索之一。以往相关研究大多关注事实性知识领域，发现4岁的儿童能够利用信息充分性线索评估信息源。在这些研究中，事实性信息的充分性线索明显，通过直接的感知觉证据即可获得。但是，学前期的儿童已经从询问事实性问题转向询问因果知识问题。运用有关因果知识的充分性线索进行选择性学习，是更加复杂的过程。它需要儿童收集客观世界中的数据推断因果关系，还需要儿童持续评估、权衡证据和陈词之间的关系，从而选择可靠的学习对象。因此，已有研究结果不能预测幼儿面对因果问题时的表现。目前为止，学前儿童是否能利用解释充分性进行选择性学习仍未知。信息源提供的解释是学前儿童获取因果知识的重要途经。但是，成人常常向儿童提供不充分的解释，可能使得儿童认为单一前因即可产生结果。鉴别不充分的解释对儿童应对复杂的学习环境至关重要。因此，本文计划通过三个研究考察在因果知识领域，学前儿童能否利用解释充分性评估陈词、选择信息源；同时，在双过程理论的基础上，考察幼儿是利用启发式策略还是利用推断策略选择信息源，及其幼儿对信息源的选择是否基于他们对他人知识的推断过程。

研究一选取249名3~5岁的学前儿童，采用“blicket检测器”任务和“冲突信息源”任务，考察学前儿童如何利用解释充分性评估陈词的质量、如何选择可靠信息源。研究一中幼儿同时接收充分和不充分两种陈词，此时，幼儿对信息源的选择可能依赖两种认知策略：启发式策略和推断策略。研究二选取60名4~5岁的学前儿童，基于双过程理论，考察幼儿选择性学习的策略是基于启发式的还是基于推断过程的。研究三在研究二的基础上，选取120名4~6岁的学前儿童，考察幼儿对信息源的推断是否基于对他人知识的推断。本研究结果表明：

（1） 学前儿童（3~5岁）能鉴别解释的充分性，他们对充分解释的评分高于不充分和错误的解释。但是，4~5岁的儿童才能区分不充分和错误的解释，并且能够自发利用解释充分性线索选择可靠的学习对象，3岁儿童只在与错误信息源同时对比时才会选择充分的信息源。

（2） 虽然4~5岁的儿童能利用充分性线索选择信息源，但是4岁儿童的选择是基于启发式策略，根据他人行为的效价（信息充分/信息不充分）对信息源做出整体判断，5岁儿童的选择则是基于对信息源内部心理状态的推断。

（3） 在选择性学习过程中，5岁儿童的选择基于对信息源知识的推断， 他们还能认识到过度自信是不合理的。年长的6岁儿童则能够进一步认识到谦逊和过度自信反映了信息源的认识水平，并选择向谦逊的信息源学习。

本研究考察了在社会学习过程中，学前期儿童对认识性线索（充分性）的利用情况，选择学习对象时采用的策略，以及选择过程中涉及的认知过程，获得了学前儿童选择性社会学习的新证据。在双过程理论的基础上，验证了学前儿童的选择性学习涉及启发式和推断过程，并揭示了其中的发展规律。这不仅丰富了理论研究，深化了研究者对学前儿童社会学习及其社会认知发展的认识，也为教育工作者因时施教、向幼儿提供充足的知识提供了理论支持。

People live in a causally complex and rich world. There are a variety of causal phenomena that children encounter with age. Fortunately, children can rely on two important information sources to acquire causal knowledge: the statistical evidence that children can directly observe and the verbal testimony about causal relations that children may receive from others. To identify novel causal relations that cannot be directly observed, children must rely on others' testimony. Previous studies on selective social learning demonstrate that children are not credulous of all the information they receive but selectively choose to learn knowledge from reliable informants. Children obtain physical evidence about causal relations through their experiences and utilize multiple social inference cues to gauge the credibility of the informants.

Most of earlier works focus on children's selective learning in the domain of factual knowledge, like the labels, functions, or locations of objects. However, during preschool age, children shift from asking "what" and "where" questions (i.e. questions about factual knowledge) to asking "why" and "how" questions (i.e. questions about causal knowledge). The evidence of the informativeness of others' factual testimony is relatively obvious (usually can be perceived directly by children). Nevertheless, the evidence of the informativeness of others' explanations must be inferred according to children's own understanding of causal relations. This difference makes it a more challenging task for children to evaluate the informativeness of others' explanations. Little is known about how children evaluate the informativeness of others' testimony about causal relations. Others' explanations of causal events play a critical role in shaping children's early learning. However, adults do not always provide high-quality explanations. Moreover, adults are biased towards expecting disjunctive relationships (in which a single factor can cause the effect alone) and learn these relationships more easily than conjunctive relationships. Taken together, adults tend to provide children with under-informative explanations conveying that one causal factor is enough to cause the effect. In this case, young children's sensitivity to others' under-informative testimony is quite essential for their learning and decision making.

In study 1，we modified the "bucket detector" paradigm and the conflicting sources paradigm to investigate whether and when 3一to 5-year-olds (N=249) utilize the informativeness of others' explanations to evaluate the explanations and choose reliable informants. Based on dual-process account, study 2 assessed the strategy that 4- and 5-year-olds (N=60) adopt in selective learning. In study 3, we examined the cognitive foundations of 4- to 6-year-olds' (N=120) selective learning, that is, whether preschoolers' selective learning was based on the general inferences about others' traits or on the specific inferences about others. The results of the current study revealed that:

a) The 3- to 5-year-olds rated the informative explanations higher than the under-informative and the inaccurate ones. However, only the 4- and 5-year-olds can distinguish the under-informative explanations from inaccurate ones. Moreover, the 4- and 5-year-olds (but not the 3-year-olds) use the informativeness of the explanations to choose reliable informants.

b) The 4-year-olds made global judgments about the informants based on the heuristic strategy, but the 5-year-olds adopt deliberative strategy to learn from more knowledgeable informants.

c) The 5-year-olds' selective learning were based on the inferences about others' knowledge. They could also recognize the implausibility of others' over-confidence. Moreover, the 6-year-olds can use the level of others' knowledge to learn from the intellectual humble informant.

Taken together, the current study investigated preschoolers' ability to use the informativeness of explanations in selective social learning. We also assessed the strategy and the cognitive foundations of children's selective learning. The findings of the current study support the dual-process account and verify that children's selective learning is based on the heuristic processes and the deliberative processes.