This paper aims to introduce a novel AI learning progression for upper-elementary students and aligns assessment items across levels of each construct to gather evidence of understanding. It also validates this quantitative measure by examining these items as two subscales for psychometric properties using the Rasch model.
Conducting a cognitive analysis of diverse data sources, including the AI4K12 big ideas (Touretzky et al., 2019), student performance on assessment items, and classroom activities from prior implementations of an AI curriculum intervention (Glazewski et al., 2022), and drawing insights from subject matter experts, this paper outlines the design of the learning progression. The second section delves into the refinement and mapping of assessment items and an evaluation of their psychometric properties to ensure the reliable placement of students within the progression.
This project identified key starting points for students and outlined how their understanding of core AI concepts should develop. The validation of the two subscales resulted in a reliable tool for accurately assessing students’ AI abilities. This tool helps educators match assessment questions to students’ current understanding and guide their progression through the learning journey.
This learning progression offers a unique framework for teaching AI to younger students, addressing a gap in K-12 education. It provides a roadmap for progressively teaching AI concepts, allowing educators to design lessons and assessments that are appropriate for students’ developmental stages.
