The state updated its sex education guidelines last year for students from pre-K …
21 NYC Schools Utilize Artificial Intelligence to Enhance Teacher Coaching
A consortium of small public high schools in New York City is examining the potential of artificial intelligence to revolutionize how teachers receive feedback on their classroom instruction.
Urban Assembly, a network of 21 schools, is collaborating with the American Institutes of Research to create an AI-powered tool that can assist instructional coaches in analyzing videos of teachers delivering lessons and providing feedback, according to network leaders.
Artificial intelligence has already revolutionized the classroom experience for many students in New York City. Students have reported that chatbots like ChatGPT have helped them comprehend difficult topics and accelerate their research. However, the implementation of this technology has faced opposition from educators and officials concerned about potential cheating, misinformation, and bias.
After initially prohibiting the use of ChatGPT on school devices due to academic integrity concerns, the New York City Education Department has committed to teaching students how to responsibly utilize the technology. They also plan to establish an institute that will explore its applications in schools.
Adopting AI for teacher coaching raises similar questions. Advocates argue that it could save instructional coaches a significant amount of time and increase access to feedback that enhances teaching quality. However, some teachers express reservations about the accuracy of the technology in capturing subtle classroom interactions, the usefulness of the data it provides, and the potential for biases.
Judy Cappuccio, a math teacher and instructional coach at Urban Assembly Institute of Math and Science for Young Women, remains open to the idea of AI assistance but maintains a healthy dose of skepticism. She believes that witnessing the tool in action and verifying its effectiveness are essential before trusting it completely.
Several schools within Urban Assembly’s network are already participating in a pilot program where educators record their teaching sessions and meticulously analyze the videos with instructional coaches to enhance their practice. Urban Assembly CEO David Adams likens this practice to athletes reviewing game footage. However, the time-consuming nature of video reviews limits the scalability of the program, resulting in insufficient feedback for teachers and infrequent evaluations.
This is where the new AI-powered tool comes into play. Over the course of a two-year rollout, researchers from the American Institutes of Research aim to develop a tool capable of measuring various aspects, such as the frequency of student-teacher interactions, laughter, and raised voices. Urban Assembly received a proposal outlining these objectives.
Initially, the tool will be implemented in the 21 schools within the Urban Assembly network. However, Adams hopes to expand its usage in the future. The estimated cost for developing, testing, and implementing the tool over two years is approximately $500,000, according to the network.
The tool will utilize natural language processing, a branch of AI that seeks to comprehend language meaning, to evaluate the tone and quality of a teacher’s language, including whether it is positive, respectful, or insulting.
Although capturing seemingly minor details through the AI tool may appear insignificant, these details can provide valuable insights into the classroom climate, from which teachers can learn. According to Adams, moments of laughter and camaraderie between students and teachers indicate that they are on the same emotional wavelength, contributing to better lesson absorption. By documenting and analyzing these moments through video recordings, teachers can replicate and foster such atmosphere.
While the tool will not replace instructional coaches, it will save them time by directing them to relevant sections of the video, generating audio transcripts, and quickly compiling data that would normally take hours for humans to gather. Adams asserts that, ultimately, the tool could expand the program’s reach and benefit more teachers.
Adams emphasizes that the tool will not serve an evaluative function or affect performance reviews conducted by school principals. Its purpose is to provide support and highlight teachers’ strengths, not solely areas for improvement.
Project leaders suggest utilizing the tool to enhance an existing teacher feedback program called CLASS, which deploys instructional coaches to evaluate educators based on video recordings of their classroom lessons. Using videos instead of live observations provides a more transparent view of the classroom and gives teachers the opportunity to observe their own teaching methods in action.
Teachers Display Interest in AI Proposal, Yet Express Concerns
Liza Backman, a science teacher and instructional coach at Urban Assembly Institute of Math and Science for Young Women in downtown Brooklyn, acknowledges the potential benefits of having quick access to the data that the AI tool can collect.
However, Backman highlights the importance of having an instructional coach to help interpret the data within the context of the class. She points out that certain lessons may not elicit laughter due to their serious nature.
Backman also raises questions about the types of school environments featured in the training videos used for the AI tool and whether biases may be inadvertently incorporated.
Adams assures that the tool being developed for Urban Assembly schools will be trained using videos from other Urban Assembly schools with similar demographics.
Other potential downsides exist as well. Soares, the principal of Urban Assembly Institute, suggests that the AI-powered tool may overlook significant moments in classroom videos that do not fit neatly into predefined categories it seeks to track. However, she believes this is a worthwhile tradeoff if it allows for the expansion of teacher participation in the program.
While some moments may go unnoticed, the benefits will extend to more individuals, Soares contends.