SBIR Phase I: Automated Performance Assessment Tools

9860581新的教育改革要求具有挑战性的标准，强调现实世界解决问题的能力，这导致了评估的范式转变，以更真实的形式支持这些技能的衡量。在真实评估运动中，一个重要的推动力是绩效评估；即基于学生完成实际任务的评估。这个小型企业创新研究第一阶段项目来自研究发展公司，旨在开发一个自动化系统，用于展示和评分性能任务。该系统将使用人工智能作为手段，使评分更及时、更经济、更可靠。研究发展公司提出，在大规模评估打分过程中，固有的主观评分过程存在劳动强度、成本和不一致性等问题，这些问题可以通过在模拟问题环境中创建绩效任务，并使用智能辅导系统和评估技术来自动评估学生来解决。具体而言，该项目将利用模拟科学实验作为绩效任务，为高中科学创造一个自动绩效评估系统，并利用相同的评分标准，将技术评分与训练有素的教师评分进行比较，从而评估技术的可靠性。研究发展公司提出了一种创新的自动化工具，用于学生的快速评估，这对学生的评估、学习过程和科学教学方式有潜在的重大影响。能够轻松衡量知识应用和综合方面的收益，对于促进我国学校更高层次思维和内容应用的价值至关重要。

9860581 New educational reforms calling for challenging standards that emphasize real world problem solving skills have caused a paradigm shift in assessment to more authentic forms that support measurement of these skills. Within the authentic assessment movement, an important thrust is performance assessment; namely assessment based on students performing realistic tasks. This Small Business Innovation Research Phase I project from Research Development Corporation seeks to develop an automated system for presenting and scoring performance tasks. The system will use artificial intelligence as a means to make scoring more timely, cost effective, and reliable. Research Development Corporation proposes that the problems of labor intensiveness, costs, and inconsistencies across an inherently subjective rating process in scoring large scale assessments can be addressed by creating performance tasks within a simulated problem environment and using intelligent tutoring system and assessment technologies to automatically assess the students. Specifically, the project will generate an automated performance assessment system for high school science using simulated science experiments as performance tasks, and it will evaluate the reliability of the technology by comparing ratings made by the technology to ratings made by trained teachers using the same scoring rubrics. Research Development Corporation proposes an innovative automated tool for the rapid assessment of students that has potential for a significant impact on student assessment, the learning process, and on the way science is taught. Being able to easily measure gains in the applications and synthesis of knowledge is crucial to promoting the value of higher level thinking and application of content in our Nation's schools.