Project I: Definition, Classification, and Risk Prediction

摘要 项目1 项目一：风险的定义、分类和预测 单词级阅读障碍（即，阅读障碍）和特殊阅读理解障碍（SRCD）是两个 重要的公共卫生问题，阅读的患病率估计在3%到20%之间 残疾和8%至10%的SRCD。该项目的长期目标是大幅增加 关于这些学习障碍的性质的可复制的知识，并在工具中实现这些知识 这可能会改善有学习障碍的个人及其家庭的结果。现有 优先考虑单个指标的阅读障碍的定义（例如，解码不佳，对 指导/干预）显示出较差一致性和纵向稳定性水平。然而，一个业务 从阅读障碍的多变量模型中得出的定义显示出明显更好的性能， 结合多个指标。具体目标1是实施阅读障碍的多变量模型作为工具 它可以在个人层面上用于预测风险，帮助识别和估计概率 关于功能上重要的结果，如使用辅助技术的可能价值。基于模型 元分析、模拟和对新数据的应用将用于生成和测试预测模型， 包括来自人工智能和贝叶斯推理的模型。具体目标2： 阅读障碍的神经生物学和行为领先指标，可能具有预测风险的价值， 识别，或预测功能显著的结果。虽然建立了关系， 基于大脑的结构（结构和功能）以及语言和识字结构，它在很大程度上是 不知道基于大脑的结构是最好的概念化的原因，后果，或仅仅是 语言和读写能力结构的相关性。潜在变化分数建模，动态系统的一种形式 建模，是一个国家的科学方法来测试替代模型，领先，滞后，或没有直接 两个结构之间的关系超越了它们仅仅相关的发展。具体目标3是进一步 了解特定阅读理解障碍的性质以及如何最好地预测 并鉴定我们打算进一步探讨这一现象的性质， 在具体目标1中已经概述了开发用于预测特定阅读理解风险的模型 残疾。最后，具体目标4是尽可能招募、研究、分析和报告分类结果 对于历史上研究不足和服务不足的人群（例如，英语学习者，居住在 贫穷、男女、种族/民族认同），同时努力实现上述三个具体目标。

ABSTRACT Project 1 Project I: Definition, Classification, and Prediction of Risk Word-level reading disability (i.e., dyslexia) and specific reading comprehension disability (SRCD) are two important public health problems, with estimates of prevalence ranging from 3 to 20 percent for reading disability and 8 to 10 percent for SRCD. The long-term objective of this project is to substantially increase replicable knowledge about the nature of these learning disabilities and to implement this knowledge in tools that potentially can improve the outcomes of individuals with learning disabilities and their families. Existing definitions of reading disability that prioritize a single indicator (e.g., poor decoding, inadequate response to instruction/intervention) show poor levels of agreement and longitudinal stability. However, an operational definition derived from a multivariate model of reading disability shows substantially better performance by combining multiple indicators. Specific aim 1 is to implement a multivariate model of reading disability as a tool that can be used at the level of the individual to predict risk, aid in identification, and estimate probabilities about functionally-significant outcomes such as the likely value of using assistive technology. Model-based meta-analysis, simulation, and application to new data will be used to generate and test prediction models, including models derived from artificial intelligence and Bayesian inference. Specific aim 2 is to identify neurobiological and behavioral leading indicators of dyslexia that may have value for predicting risk, aiding identification, or in predicting functionally significant outcomes. Although established relations exist between brain-based constructs (both structural and functional) and language and literacy constructs, it is largely unknown whether the brain-based constructs are best conceptualized as causes, consequences, or mere correlates of the language and literacy constructs. Latent change score modeling, a form of dynamic systems modeling, is a state-of-the-science approach to test alternative models that posit leading, lagging, or no direct relations between two constructs beyond their mere correlated development. Specific aim 3 is to further understanding about the nature of specific reading comprehension disability and how it best can be predicted and identified. We intend to further explore the nature of this phenomenon and to use the approach we will have outlined in specific aim 1 to develop a model for predicting risk of specific reading comprehension disability. Finally specific aim 4 is to recruit, study, analyze, and report disaggregated results where possible for historically understudied and underserved populations (e.g., English Language Learners, families living in poverty, males and females, racial/ethnic identity) as we work to achieve the previous three specific aims.