Traditional assessment methods struggle to authentically capture the practical application of competencies in complex learning environments such as SEL (Social-Emotional Learning) and CTE (Career and Technical Education). These fields require tools that go beyond self-reports or structured written assessments, which often miss the nuanced, context-driven ways students demonstrate skills. ZoneSight leverages generative AI to perform deep contextual analysis on unstructured data like audio and video, allowing for a richer, more flexible assessment of competencies that more closely reflect the authentic learning, while also introducing more objectivity and scalability to assessments in these fields. This paper presents early progress and preliminary results from ZoneSight, a tool designed to address this challenge by analyzing real interactions to extract insights into nuanced competencies, beginning with SEL. ZoneSight's scalable, psychometrically sound assessments align with the dynamic nature of SEL (current focus) and hold future potential for application in CTE environments. © 2025 IEEE.
Author Keywords
Index Keywords
