Human Tutoring Augmented by Artificial Intelligence (AI): A Tutoring Analytics and Performance Support Model to Improve the Work and Professional Growth of Future Tutors

这项研究项目解决了国家对大幅扩大辅导队伍的迫切需要，辅导队伍主要由受过有限正规教育培训的准专业人员组成。人类辅导已经多次被证明对学生的学习产生了巨大的影响，现在正准备成为日常学校教育的核心结构部分。导师可以在解决不断扩大的成绩和机会差距以及照顾学生日益增长的社会情感需求方面发挥重要作用。随着辅导队伍的扩大，新技术和专业学习模式对于维护和提高辅导对学生的学习益处至关重要，其中包括来自历史上被边缘化群体和低社会经济背景的学生。该项目将开发一种用于在职培训的人工智能增强的专业支持模型，该模型以个性化的数据驱动反馈为中心，以增强导师的实践以及他们的专业成长。该项目将为培养教师向所有学生提供高质量教学的能力提供重要的新知识，并使辅导成为进入教育专业的一条途径，潜在地有助于解决全国严重缺乏合格教师的问题。该项目汇集了科罗拉多大学博尔德大学的一个跨学科研究团队，与佐贺教育的从业者合作伙伴合作。前者在语音和语言处理的深度学习、人机交互、数学教育、教师学习和指导、教育政策和行为科学方面拥有专业知识。作为对这些观点的补充，后者在开发、组织和支持第一标题公立学校的数学辅导计划方面提供了丰富的经验。这项研究的具体工作背景将是佐贺教育提供的现有的在校数学辅导计划。该项目将开发和研究一个创新的辅导分析系统，以及一个支持教师在日常工作中学习有效教学技能的工具包。该系统将通过对教师在与学生合作时使用的话语策略提供近乎实时的反馈，来增强教师的教学和关系技能。更具体地说，深度学习模型将用于自动分析教师与学生的对话并识别重要的话语事件。多模式反馈界面将为教师提供与每个辅导课程相关的个性化信息。这种多层次的方法还将包括支持教师和监督教练在专业学习期间使用反馈进行反思注意和设定目标的资源。研究将调查这种模式在多大程度上改变了工作的性质(辅导互动和实践)，并导致了工人的职业化(导师的效率和工作敬业度)。该项目由人类-技术前沿的未来工作交叉发展计划资助，旨在通过推进与人类工人和谐运作的智能工作技术的设计，促进对工作环境中相互依赖的人-技术伙伴关系的更深层次的基本理解。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

This research project addresses the urgent national need to dramatically enlarge the tutoring workforce, which is primarily comprised of paraprofessionals with limited formal training as educators. Human tutoring has been repeatedly shown to produce large impacts on student learning, and is now poised to become a core structural part of daily schooling. Tutors can play an important role in addressing widening achievement and opportunity gaps as well as in attending to students’ increased social-emotional needs. As the tutoring workforce expands, new technologies and professional learning models will be critical to preserve and enhance tutoring's learning benefits for students, including students from historically marginalized groups and low socio-economic backgrounds. This project will develop an AI-augmented professional support model for on-the-job training that centers personalized, data-driven feedback in order to enhance tutors’ practices as well as their professional growth. This project will contribute vital new knowledge towards building capacity for tutors to provide high quality instruction to all students and for tutoring to serve as a pathway into education professions, potentially helping to address the dire national shortage of qualified teachers.This project brings together an interdisciplinary team of researchers from the University of Colorado Boulder working with practitioner partners from Saga Education. The former group has expertise in deep learning for speech and language processing, human-computer interaction, mathematics education, teacher learning and coaching, educational policy, and behavioral science. Complementing these perspectives, the latter group offers extensive experience in developing, organizing, and supporting mathematics tutoring programs in Title I public schools. The specific work context for this research will be an existing in-school, mathematics tutoring program provided by Saga Education. The project will develop and study an innovative tutoring analytics system, along with a toolkit that supports the learning of effective pedagogical skills as tutors engage in their everyday work. The system will enhance tutors’ pedagogical and relational skills by providing near-real-time feedback on the discourse strategies they use when working with students. More specifically, deep learning models will be used to automatically analyze tutor-student dialogues and recognize important discursive events. Multimodal feedback interfaces will provide tutors with personalized information related to each tutoring session. This multilayered approach will also include resources that support tutors’ and supervising coaches’ use of the feedback for reflective noticing and goal-setting during professional learning sessions. Research will investigate the extent to which this model changes the nature of work (tutoring interactions and practices) and leads to the professionalization of workers (tutor efficacy and work engagement).This project has been funded by the Future of Work at the Human-Technology Frontier cross-directorate program to promote deeper basic understanding of the interdependent human-technology partnership in work contexts by advancing the design of intelligent work technologies that operate in harmony with human workers.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.