The role of parent and family factors in child language development

描述（由申请人提供）：拟议的研究重点是开发早期儿童语言发展和家长投入的新措施，并使用这些措施来预测儿童在幼儿园的识字技能。该项目的指导阶段的目的是：1）接受测量模型的培训，特别是项目反应理论，2）将这些模型应用于儿童词汇数据，以开发新的测量方法，重点关注产出的稀有性而不是频率，3）验证这些测量方法，以及4）通过分层线性模型（HLM）的培训来模拟这些新测量方法的增长。独立阶段的重点是使用在指导阶段学到的技能：1）应用类似的模型来开发和验证新的儿童句法和父母输入的措施，2）检查父母输入措施和儿童词汇和句法增长之间的关系，3）使用新的儿童语言和父母输入措施作为儿童识字技能的预测因子。这些数据来自两个现有的亲子互动成绩单数据集。第一个数据集来自于美国大学的语言发展项目。 芝加哥，包括90分钟的互动，每4个月之间的儿童年龄14-54个月， 来自不同背景的60个家庭的样本。第二个数据集来自哈佛大学的Early Head Start研究，包括108个低收入家庭样本中14-36个月的儿童之间每年一次的10分钟互动。这两个样本都有额外的数据，使用家长报告和标准化的措施，儿童语言验证新的措施，这两个样本的措施，儿童的识字能力在幼儿园使用的最后阶段的研究。研究计划的总体目标是：（1）开发儿童早期词汇和句法技能的测量方法，以捕捉产出的稀有性，（2）更好地了解特定的父母输入措施和儿童语言发展之间的关系，（3）进一步研究早期口语能力在预测后期识字技能中的作用。相关性：早期语言技能是识字技能和学校成功的有力预测因素。这项研究将使我们更深入地了解儿童的早期语言能力，父母的投入在儿童语言发展中的作用，以及儿童的早期语言和后来的识字技能之间的具体关系。重点是低收入儿童，那些最有可能学业失败的儿童。

DESCRIPTION (provided by applicant): The proposed research focuses on developing new measures of early child language development and parent input, and using those measures to predict children's literacy skills in kindergarten. The purpose of the mentored phase of the project is to: 1) receive training in measurement models, specifically Item Response Theory, 2) to apply these models to child vocabulary data to develop new measures that focus on rarity of production rather than frequency, 3) to validate these measures, and 4) to model growth of these new measures via training in Hierarchical Linear Modeling (HLM). The independent phase focuses on using the skills learned during the mentored phase to: 1) apply similar models to develop and validate new measures of child syntax and parent input, 2) examine the relation between parent input measures and child vocabulary and syntactic growth, and 3) use the new child language and parent input measures as predictors of child literacy skills in kindergarten. The data come from two existing datasets of transcripts from parent-child interaction. The first dataset is from the Language Development Project at the University of Chicago and consists of 90-minute interactions every 4 months between child ages 14-54 months in a sample of 60 families from diverse backgrounds. The second dataset is from the Early Head Start study at Harvard University and consists of 10-minute interactions once a year between child ages 14-36 months in a sample of 108 low-income families. Both samples have additional data using parent report and standardized measures of child language for validating new measures, and both samples have measures of children's literacy skills at kindergarten for use in the final phase of the research. The overall goals of the research program are: i) to develop measures of children's vocabulary and syntactic skills during early childhood that capture rarity of production, 2) to better understand the relation between specific parent input measures and child language growth, and 3) to further the research on the role of early oral language abilities in predicting later literacy skills. Relevance: Early language skills are a strong predictor of literacy skills and school success. This research will provide us with deeper understanding of children's early language abilities, of the role of parental input in child language development, and of the specific relations between children's early language and later literacy skills. The focus is on low-income children, those most at risk for school failure.