Using Classical Item Analysis to Validate Learning Paths in an Online Course
Abstract: This study compares the instructional sequences defined by experts for an online course with the learning paths indicated by measurement data. The goal was to determine which of the expert-identified topic sequences were supported by total population statistics and which were not. In the cases where the expert sequence was not supported, relationships were modified through expert judgment to incrementally discover a sequence (precedence structure) that was supported both by experts and by the measurement data for the total population. Additional side benefits to item alignment quality and grouping of skills within curriculum will be discussed.
Presider: Sarah Haavind, Lesley University