Mining & Modeling Backward & Depth-First Processes Used in Diagramming Complex Arguments

• In Event: Learning Analytics

Short Description

Research on the efficacy of argument diagramming reveal mixed results because the precise processes used to construct diagrams has yet to be fully operationalized and measured in ways that enable diagramming tools to assess and assist students in real-time. This study mined students’ diagramming actions, developed algorithms to operationalize and measure backward, forward, breadth, depth-first reasoning, compared regression models to identify algorithms that produce measures that best predict diagram scores, and determined the relative impact of each process on diagram scores.

Author

• Presenter: Allan Jeong, Florida State University