Effectiveness and predictors of response for a technology-based reading intervention in the home

项目概要/摘要。 阅读障碍（RD）是最常见的学习障碍，影响10 - 15%的学龄儿童。它 在生命的各个阶段都会出现严重的功能障碍，大量数据记录了终身的不利因素 教育和职业成就。因此，确定有效和负担得起的RD治疗方法是 阅读研究人员、临床医生和教育工作者的高度优先事项。有问题的是，目前的循证阅读 干预措施在很大程度上依赖于训练有素的专家的服务，无论是在资源充足的教室还是在临床 设置.因此，资源不足的学校（或国家）往往无法为儿童提供阅读干预。 他们的学生近年来，基于技术的阅读干预被提出作为一种手段， 克服这些挑战，因为它们可以在家里管理，没有直接的专家监督- 从而最小化资源需求。在以阅读为中心的教育技术领域，GraphoLearn已经成为一个 领导者，拥有迄今为止最大的证据基础。然而，迄今为止，绝大多数研究都是在 在高度受控的环境中，而不是在家庭环境中，这是为了-留下开放的问题 关于效率。此外，与任何干预措施类似，并非所有RD儿童都能从以下方面平等受益： 治疗;然而，很少注意识别治疗反应的预测因子。 在这里，我们建议通过实施来评估家庭管理的GraphoLearn的有效性 在450名阅读障碍儿童（男孩和女孩，年龄6.0- 10.0）。为了以最小的成本快速实现这一目标，我们将从健康大脑中招募参与者。 网络[HBN]，一项正在进行的关于5.0-21.0岁儿童心理健康和学习障碍的研究， 家庭对行为和/或学习有担忧（目标n = 10，000;当前n = 3000+）。的可用性 全面的表征（例如，人口统计学、认知、心理健康、EEG、多模态MRI） 参与者使样本最适合探索广泛的参与者和环境因素 这可能影响治疗结果（即，人口统计学、认知、情感、神经生物学、环境）。 建议的工作的具体目标是：1）在大样本中评估GraphoLearn的有效性， RD儿童，和2）确定参与者相关的和环境水平的因素， 与GraphoLearn结果相关。为了实现后一个目标，复杂的机器学习 将采用随机森林回归模型（Random Forest Regression Models）。

Project Summary/Abstract. Reading Disability (RD) is the most common learning disability, affecting 10 – 15% of school age children. It incurs major functional impairments at all stages of life, with a wealth of data documenting lifelong disadvantages in educational and occupational attainment. Therefore, identifying effective and affordable treatments for RD is a high priority for reading researchers, clinicians and educators. Problematically, current evidence-based reading interventions largely rely on services by trained specialists, either in well-resourced classrooms or clinical settings. As such, under-resourced schools (or countries) often are unable to provide reading interventions for their students. In recent years, technology-based reading interventions have been proposed as a means of overcoming these challenges, as they can be administered in the home, without direct expert supervision - thereby minimizing resource demands. In the area of reading-focused EdTech, GraphoLearn has emerged as a leader, with the largest evidence-base to date. However, the vast majority of studies to date have been conducted in highly controlled settings, rather than the home environment it was intended for – leaving open questions about effectiveness. Additionally, similar to any intervention, not all children with RD benefit equally from treatment; however little attention has been given to identifying predictors of treatment response. Here we propose to evaluate the effectiveness of home-administered GraphoLearn through the implementation of a large-scale, randomized controlled trial (RCT) in 450 reading disabled children (boys and girls, ages 6.0- 10.0). To accomplish this goal rapidly and with minimal cost, we will recruit participants from the Healthy Brain Network [HBN], an ongoing study of mental health and learning disorders in children, ages 5.0-21.0, whose family have concerns about behavior and/or learning (target n = 10,000; current n = 3000+). The availability of comprehensive characterizations (e.g., demographic, cognitive, mental health, EEG, multimodal MRI) for all HBN participants makes the sample optimal for exploring an extensive set of participant and environmental factors that may affect treatment outcomes (i.e., demographic, cognitive, emotional, neurobiological, environmental). Specific aims of the proposed work are to: 1) Evaluate the effectiveness of GraphoLearn in a large sample of children with RD, and 2) Identify participant-related and environment-level factors that are significantly associated with GraphoLearn outcomes. To accomplish this latter aim, sophisticated machine learning approaches (Random Forest Regression models) will be employed.