Constraints on Visual Statistical Learning in Infancy

婴儿期视觉统计学习的限制

• 批准号: 8690126
• 负责人: SCOTT P JOHNSON
• 金额: $ 30.39万
• 依托单位: UNIVERSITY OF CALIFORNIA LOS ANGELES
• 依托单位国家: 美国
• 财政年份: 2012
• 资助国家: 美国
• 起止时间: 2012-07-20 至 2017-05-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/8690126
• 关键词: Accounting; Achievement; Address; Affect; Age; Appearance; Attention; Auditory; Autistic Disorder; Child; Cognitive; Complex; Cues; Detection; Development; Developmental Delay Disorders; Dimensions; Discrimination; Early Diagnosis; Early treatment; Environment; Eye; Eye Movements; Goals; Human; Infant; Iron deficiency anemia; Knowledge; Language Development; Lead; Learning; Light; Linguistics; Machine Learning; Memory; Methods; Modeling; Older Population; Pattern; Perception; Performance; Population; Probability; Procedures; Process; Research; Risk; Role; Scanning; Series; Shapes; Specificity; Stimulus; Structure; Testing; Time; Ursidae Family; Visual; Work; autism spectrum disorder; coping; developmental disease; disability; infancy; interest; novel; research study; skills; statistics; theories; tool; trend; visual information; visual stimulus

DESCRIPTION (provided by applicant): Visual statistical learning is the process of identifying patterns of probabilistic co-occurrence among visual features, essential to our ability to perceivethe world as predictable and stable. The proposed experiments will provide the first comprehensive examination of infants' detection of complex statistical patterns in visual sequences and layouts b describing computation of probabilistic information by infants from 2 to 14 months. This is a formaive time in perceptual and cognitive development characterized by rapid developmental change in perception and learning of environmental structure. In particular, the experiments will examine how structural variability, cognitive load, and memory limitations affect learning, how contextual cues facilitate or impede this learning, and whether the products of such learning can be generalized toa different setting. In addition, the experiments will provide critical tests of domain-specificity b examining the specific contributions of spatial information to visual statistical learning. The shot-term objectives of the proposed research are to learn how developing perceptual and cognitive skills intract in early development to identify statistically defined patterns. The long-term goals are to clarifytheories of cognitive development that posit a central role for inductive learning by computing probabilitie of observations. The results of this research may have implications for understanding development in children who may be at risk for developmental disorders such as iron-deficiency anemia and autism spectrum disorders, both of which have been characterized as deficits in implicit learning. Such an understanding may lead to assessment tools more closely tailored to early diagnosis and treatment than are presently available.

描述(申请人提供)：视觉统计学习是识别视觉特征之间的概率共现模式的过程，对于我们感知世界是可预测的和稳定的能力至关重要。拟议的实验将首次全面检查婴儿对视觉序列中复杂统计模式的检测，以及描述2至14个月婴儿计算概率信息的布局b。这是知觉和认知发展的一个形成时期，其特征是对环境结构的知觉和学习的快速发展变化。特别是，这些实验将检验如何 结构可变性、认知负荷和记忆限制影响学习，以及背景线索如何 促进或阻碍这种学习，以及这种学习的产品是否可以推广到不同的环境。此外，这些实验将提供领域特异性的关键测试，以检验空间信息对视觉统计学习的具体贡献。拟议研究的短期目标是学习如何在早期发展中发展感知和认知技能，以识别统计定义的模式。长期目标是澄清认知发展的理论，这些理论通过计算观察的概率来确定归纳学习的核心作用。这项研究的结果可能对理解儿童的发育有意义，这些儿童可能有患发育障碍的风险，如缺铁性贫血和自闭症谱系障碍，这两种疾病都被描述为内隐学习缺陷。这样的一种 了解可能导致评估工具比目前可用的更紧密地为早期诊断和治疗量身定做。