Developing Dynamic Forecasting Intervention Models for Upper Elementary Students

为高年级小学生开发动态预测干预模型

• 批准号: 8035948
• 负责人: Carol McDonald Connor
• 金额: $ 6.73万
• 依托单位: FLORIDA STATE UNIVERSITY
• 依托单位国家: 美国
• 财政年份: 2010
• 资助国家: 美国
• 起止时间: 2010-03-01 至 2012-06-30
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/8035948
• 关键词: Accounting; Affect; Algorithms; Area; Characteristics; Child; Child Language; Child health care; Code; Cognitive; Complex; Comprehension; Coupled; Data; Development; Education; Exploratory/Developmental Grant; Eye; Funding; Growth; Human Development; Individual; Institutes; Instruction; Intervention; Intervention Studies; Knowledge; Language; Language Arts; Lead; Learning; Linear Models; Link; Methodology; Methods; Modeling; Nature; Pattern; Reader; Reading; Research; Research Personnel; Research Project Grants; Role; Schools; Social Sciences; Students; Study Section; Testing; Text; base; design; elementary school; falls; fifth grade; innovation; literacy; meetings; multilevel analysis; public health relevance; science education; skills

DESCRIPTION (provided by applicant): The proposed study has several specific, interrelated aims: (1) To closely examine the nature and variability of the classroom instruction 4th-5th grade children receive; (2) To investigate children's language and literacy development from 4th grade through the end of 5th grade including development of comprehension strategies and use of inferences in reading narrative and expository text; this includes the use of reading comprehension strategies using eye-tracking methodology; (3) To examine the contribution of 4th and 5th grade instruction to children's language and literacy development. This includes examining whether the effect of specific instructional strategies depends on individual children's language and literacy skills. That is, are there child characteristic by instruction (child X instruction) interactions? and (4) To develop dynamic forecasting intervention models for 4th and 5th grade students, which take into account child X instruction interactions. Dynamic forecasting intervention models use multilevel models to predict optimal patterns of instruction (i.e., that lead to greater reading skill growth) for individual children based on their language and literacy skills. The study aims will be met by building on and extending research that investigates the effect of individualized literacy instruction from 1st through 3rd grade on students' literacy development. Two studies in 1st grade reveal that children, who received individualized instruction, using dynamic forecasting intervention models, demonstrated greater reading skill growth than did children in control classrooms. For the proposed study, students who are currently participating in the 3rd grade study and who have completed the 1st and 2nd grade study (n = 225), and their classmates, will be followed into 4th and 5th grade (total anticipated n = 550 students, 30 4th and 5th grade classrooms in 8 schools). Their language and literacy skills will be assessed in the fall and spring. Classroom instruction will be observed, video-taped, and coded for each child late fall, winter, and early spring. Eye-tracking methods will be used to examine more closely children's use of reading strategies. Data from this study will permit the close examination of children's literacy skill development - including identifying the complex and transactional associations among key components of literacy that may respond to particular types of instruction. Multilevel hierarchical linear modeling, including cross-classified random effects models, will be used to conduct the analyses. By explicating quantifiable, albeit complex, links between children's assessed language and literacy skills and amounts and types of instruction to be provided, dynamic forecasting intervention models may have important implications for increasing understanding of how children learn complex multidimensional cognitive tasks (such as reading) and the critical role of instruction from first through fifth grade in facilitating this learning. Long term, these results have the potential of guiding the design and implementation of more effective literacy instruction in the upper elementary grades. PUBLIC HEALTH RELEVANCE: Proficient literacy skills are critical in today's information driven world and yet two-thirds of children fail to achieve proficient reading skills during elementary school. By understanding how 4th and 5th grade classroom instruction, across the content areas, affects children's literacy development, we can develop dynamic forecasting models of proficient reading that can be used to design and implement more effective instruction for 4th and 5th graders.

描述（由申请人提供）：拟议的研究有几个具体的，相互关联的目标：（1）密切审查的性质和可变性的课堂教学4 - 5年级儿童接受;（2）调查儿童的语言和识字发展从4年级到5年级结束，包括发展的理解策略和使用推理的阅读叙事和简要文本;这包括使用眼动追踪法的阅读理解策略的使用;（3）考察四、五年级教学对儿童语言和读写能力发展的贡献。这包括研究特定教学策略的效果是否取决于个别儿童的语言和识字技能。也就是说，是否存在儿童特征与指令（儿童X指令）的相互作用？（4）发展四、五年级学生的动态预测干预模式，并考虑儿童X教学互动。动态预测干预模型使用多级模型来预测最佳的教学模式（即，这导致更大的阅读技能增长）为个别儿童的基础上，他们的语言和识字技能。本研究的目的将通过建立和扩展研究，调查从1到3年级的个性化识字教学对学生的识字发展的影响。一年级的两项研究表明，儿童，接受个性化的指导，使用动态预测干预模型，表现出更大的阅读技能的增长比控制教室的儿童。对于拟议的研究，目前正在参加三年级研究并已完成一年级和二年级研究的学生（n = 225）及其同学将被跟踪到四年级和五年级（预计总人数= 550名学生，8所学校的30间四年级和五年级教室）。他们的语言和识字能力将在秋季和春季进行评估。课堂教学将被观察，录像，并为每个孩子编码秋末，冬季和早春。眼动追踪法将被用来更仔细地检查儿童的阅读策略的使用。这项研究的数据将允许仔细检查儿童的识字技能的发展-包括确定复杂的和交易的关联识字的关键组成部分，可能会响应特定类型的教学。将使用多水平分层线性模型（包括交叉分类随机效应模型）进行分析。通过阐明可量化的，虽然复杂的，儿童的评估语言和识字技能和数量和类型的指令提供之间的联系，动态预测干预模型可能有重要的影响，增加了解儿童如何学习复杂的多维认知任务（如阅读）和关键作用的指令从一年级到五年级，以促进这种学习。从长远来看，这些结果有可能指导设计和实施更有效的识字教学在小学高年级。 公共卫生相关性：在当今信息驱动的世界中，熟练的识字技能至关重要，但三分之二的儿童在小学期间未能达到熟练的阅读技能。通过了解四年级和五年级的课堂教学，跨内容领域，影响儿童的识字发展，我们可以开发动态预测模型，可以用来设计和实施更有效的教学，为四年级和五年级的熟练阅读。