EHP: Understanding the Interactions Between Starting Values and Early Interventions that Explain Children's Academic Performance

项目简介：0338740参考译文：美国佛罗里达大学NSF项目：提高人的表现PERFORMANCEPI：Figlio，David NTITLE：了解初值和早期干预之间的相互作用，以解释儿童的学习成绩在进入公立学校时，不同种族、民族和社会经济群体的学生之间存在着非常大的平均表现差距。此外，组内表现的差异远远超过正常测量的组之间的差异。这表明，种族、民族、贫困和特定于学校的因素只解释了随后学习成绩差异的一小部分。几十年来，政策制定者一直在设计健康、教育和社会福利政策，以帮助贫困家庭的儿童创造公平的竞争环境。本研究调查了小学和中学期间群体之间和群体内成绩差距的决定因素和演变，重点关注在儿童早期经历的教育、医疗和社会福利政策干预的作用。早期关于儿童早期干预的研究有几个数据问题，这些问题可能会阻碍我们理解公共政策在解释早期和中期教育结果方面的作用。这项研究将克服以前关于这一主题的研究中的主要弱点；它将出生生命统计数据与小学记录以及来自多个来源的关于幼儿参与干预的行政记录合并在一起。将考虑一系列的早期干预计划，从为新生儿重症监护毕业生服务的后续计划到Head Start，甚至Start。这些数据集非常大，非常详细，分析数据集将使用佛罗里达州的每一名学生构建，这些学生的出生记录可以与教育数据相匹配，包括特殊教育安置、出勤率、行为推荐和该州强制进行的综合评估测试的分数。大型管理数据集的使用减少了测量问题，也有助于研究不同的政策效果。大量和丰富的数据允许研究各种相互关联的问题，包括：(I)基于学校的学前教育计划是否改善了低收入家庭儿童的表现；(Ii)参与社会服务计划对儿童表现的影响。统计技术直截了当，恰如其分。一些变量是内生的，但数据足够丰富，足以让PI提供可信的识别策略来确定因果关系。这项研究的结果将对教育政策产生重要的影响，不仅在美国，而且可能在全世界。

ABSTRACTPROPOSAL NO: 0338740INSTITUTION: University of FloridaNSF PROGRAM: ENHANCING HUMAN PERFORMANCEPI: Figlio, David NTITLE: Understanding the Interactions Between Starting Values and Early Interventions that Explain Children's Academic PerformanceAt entrance to public school, there exist very large average performance gaps across students of different races, ethnicities, and socio-economic groups. Moreover, the within-group variation in performance far exceeds the differences among normatively measured groups. This suggests that race, ethnicity, and poverty, and school-specific factors, explain only a small portion of the differences in subsequent academic performance. Policymakers have for decades designed health, education, and social welfare policies to help level the playing field for children from disadvantaged backgrounds.This research investigates the determinants and evolution of performance gaps across and within groups during elementary and middle school, focusing on the roles of educational, medical, and social welfare policy interventions experienced during early childhood. Earlier studies of early childhood interventions have several data problems that may impede our understanding of the role of public policy in explaining early and mid-run educational outcomes. This research will overcome the major weaknesses in previous studies on the subject; it merges cohorts of birth vital statistics to elementary school records, and administrative records on early childhood intervention participation drawn from multiple sources. A wide array of early intervention programs, ranging from follow-up programs serving neonatal intensive care graduates to Head Start and Even Start will be considered. The data sets are extremely large and very detailed, and the analysis data set will be constructed using every student in the state of Florida whose birth records can be matched with education data, including special education placement, attendance, behavioral referrals and scores on the state's mandated comprehensive assessment test. The use of large administrative data sets reduces measurement problems, and also facilitates the study of heterogeneous policy effects.The large and rich data allows a study of a wide variety of interrelated problems, including: (i) whether school-based pre-kindergarten programs improve the performance of children from low-income families and (ii) the effects of social service program participation on children's performance. The statistical techniques are straightforward and appropriate. Some of the variables are endogenous but the data is rich enough for the PI to provide credible identification strategies to determine causal effects. The results of this research will have important educational policy implications, not only in the US but probably throughout the world.