AI-assisted coaching: A new strategy for quality at scale

• In Event: Harnessing AI for teacher empowerment: Teacher development, self-reflection, and lesson planning

Proposal

The growing emergence of new technologies is creating game-changing opportunities to advance foundational education outcomes for children around the world. One of the most important investments a country can make for the education of its children is skilled teachers (UNESCO, 2016). Coaching and mentoring support to teachers is widely recognized as a pivotal factor in effectively improving teaching practices. Without follow-up through a means such as coaching, teachers adopt only 10 percent of new skills from a workshop (Smith & Gillespie, 2007). Currently, Room to Read offers coaching to teachers to strengthen instructional practices through classroom observation, feedback, and training. This coaching model is largely low-tech and people-driven and informed by conventional methods of data collection and analysis which are often time-consuming and costly to implement. It relies on the knowledge and experience of the coach, leading to uneven results. To provide quality coaching at scale, we need to ensure that a large number of coaches have the necessary knowledge at their fingertips to analyze teaching practices and provide quality feedback to teachers.

As more industries and organizations begin to explore the applications of artificial intelligence, it is imperative to begin testing how artificial intelligence tools can enhance education-sector interventions, especially given the urgent need to address recent global learning losses and shortages of qualified teachers. Room to Read, with the support of the Jacobs Institute for Innovation in Education, launched a project that explores how AI can improve literacy coaches’ capacities to (1) track school and teacher progress; (2) provide more customized, structured coaching and targeted feedback to teachers to improve their skills; and (3) provide real-time feedback to teachers and coaches. With better information and resources, more coaches can enable teachers to strengthen practices and provide action-oriented inputs that strengthen the feedback loop between training and classroom implementation. The project leverages natural language processing and machine learning techniques via an AI-based chatbot tool. Using the new AI app, coaches will enter classroom observation data and the chatbot will generate suggestions for feedback and resources based on a custom-built knowledge base on the science of reading and Room to Read’s literacy curriculum, current and past teacher observations, and coaching plans. Future iterations will include a teacher interface where teachers can pose questions or upload video for analysis and feedback.

This presentation will explore key lessons from the development stage of the project, reflecting on the creation of the knowledge base, quality of the coaching suggestions, and process of training the AI application. It will conclude with recommendations for governments and organizations looking to scale quality coaching for teachers through AI.



Smith, C., & Gillespie, M. (2007). Research on Professional Development and Teacher Change: Implications for Adult Basic Education. In J. Comings, B. Garner, & C. Smith (Eds.), Review of Adult Learning and Literacy (Vol. 7). Routledge. https://www.ncsall.net/fileadmin/resources/ann_rev/smith-gillespie-07.pdf
UNESCO. (2016). Global Education Monitoring Report (Policy Paper 23; Every Child Should Have a Textbook: Policy Paper 23). UNESCO.

Author

• Kerri Thomsen, Room to Read