SBIR Phase I: Mapping a Pathway to College Using Predictive Analytic Modeling and Decision Support

SBIR 第一阶段：利用预测分析模型和决策支持规划大学之路

• 批准号: 1548674
• 负责人: Angie Eilers
• 金额: $ 15万
• 依托单位: UR TURN, SBC
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2016
• 资助国家: 美国
• 起止时间: 2016-01-01 至 2016-06-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1548674&HistoricalAwards=false
• 关键词: SBIR; Phase; Mapping; Pathway; College

The SBIR Phase I project supports six months of R/R & D to develop a goal-setting and planning software tool for students from 5th to 12th grade and their families, not unlike current tools for financial planning or trip planning. Students and families will benefit from the power of prediction-based goal setting as well as an early warning system that indicates when students are off track of their education goal, and evidence-based program suggestions to get students back on track of their goal. Navigating the education system in the U.S. is complex, even more so for first-generation college seekers and new immigrant students and families. Between 9th grade and the first year of college, 50% of America's students fall off the path toward higher education even as research shows that students cannot make a living wage without a post-secondary diploma. Today, nearly every U.S. school offers an electronic grade book and a parent portal that shares student performance data with students and their families. This SBIR-sponsored tool would integrate seamlessly with these existing portals to offer added-value features to a system that has not changed in over 15 years. With notoriously poor counselor-to-students ratio of 471:1, and with the uneven distribution of access to supports (such as private counseling services for those who can afford it), technology such as this may very well serve as a leveling instrument in U.S. schools. SBIR support provides the opportunity for technology transfer of what we know about financial planning and business practices to inform education planning and goal-setting based purely on data analytics and best practices.This SBIR-sponsored innovation harnesses the advances of business intelligence and decision support intelligence and applies it to education. This is a new-to-the-market tool with graphic displays of data, progress mapping, predictive analytics, research-based and customized interventions associated with improved student outcomes, all within parent portal systems that are widely used across the U.S. This web and mobile-enabled tool monitors performance and progress, not unlike what the mobile wrist band devices that offer health enthusiasts today. After initial 6 months of funding, the project will have determined proper alignment of data sets for analytic purposes, will have developed predictive analytic models using de-identified student data, will have developed a data visualization dashboard that showcases the power of this forward-looking tool on school web-based platforms, and will have conducted feedback sessions with customers on their interest level, ease of use, and accuracy of predictive outcomes. The methods used will include multiple regression, logistic regression and descriptive analysis. Prediction accuracy will be evaluated by logistic regression and ROC curve analysis.

SBIR第一阶段项目支持六个月的R/R D，为5至12年级的学生及其家庭开发一个目标设定和规划软件工具，与目前的财务规划或旅行规划工具没有什么不同。学生和家庭将受益于基于预测的目标设定的力量，以及一个早期预警系统，该系统可以指示学生何时偏离他们的教育目标，以及基于证据的计划建议，让学生回到他们的目标轨道上。在美国的教育系统中导航是复杂的，对于第一代大学寻求者和新移民学生和家庭来说更是如此。在9年级到大学一年级之间，50%的美国学生放弃了接受高等教育的道路，尽管研究表明，没有高等教育文凭的学生无法维持生计。如今，几乎每所美国学校都提供电子成绩册和家长门户网站，与学生及其家人分享学生的成绩数据。这个由SBIR赞助的工具将与这些现有的门户网站无缝集成，为一个15年来没有变化的系统提供增值功能。众所周知，辅导员与学生的比例为471：1，而且获得支持的机会分配不均（例如为那些负担得起的人提供私人咨询服务），像这样的技术很可能成为美国学校的一种平衡工具。SBIR支持为我们提供了一个机会，将我们对财务规划和业务实践的了解进行技术转让，从而为纯粹基于数据分析和最佳实践的教育规划和目标设定提供信息。这项SBIR赞助的创新利用了商业智能和决策支持智能的进步，并将其应用于教育。 这是一个新的市场工具与数据的图形显示，进度图，预测分析，基于研究和定制的干预与改善学生的成绩，所有在美国广泛使用的家长门户系统内这个网络和移动功能的工具监控性能和进度，没有什么不同的移动的腕带设备提供健康爱好者今天. 经过最初6个月的资助，该项目将确定用于分析目的的数据集的适当对齐，将使用去识别的学生数据开发预测分析模型，将开发一个数据可视化仪表板，展示这个前瞻性工具在学校基于网络的平台上的力量，并将与客户就他们的兴趣水平，易用性，和预测结果的准确性。所用方法包括多元回归、逻辑回归和描述性分析。 预测准确性将通过逻辑回归和ROC曲线分析进行评估。