Mining and Modeling the Cognitive Strategies Used to Construct Argument Versus Causal Maps in Computer-Aided Diagramming Tools

Despite four decades of research demonstrating the positive impact of computer-aided diagramming tools on student learning, there remains a lack of research that identifies the cognitive strategies used by students (and enabled by the tools) to create higher-quality maps and achieve deeper understanding. This chapter reports two studies examining students’ cognitive processes to construct argument and causal maps using the computer-aided diagramming tool jMAP. Students’ mapping actions were mined and used to develop algorithms to detect and measure students’ use of backward, forward, breadth-first, and depth-first reasoning. The first study revealed that observing the placement of the first five nodes in relation to previously moved nodes in students’ argument maps was sufficient to predict map scores and that the ratio between the use of backward versus forward and the use of breadth versus depth-first processes (not individual frequency counts) predicted map scores. The study found that students’ backward and depth-first processing correlated with higher map scores. In contrast, analysis of causal maps (using the same algorithms) showed that all reasoning processes produced maps of similar quality, with backward processing contributing significantly more to map scores than depth-first processing. These differences in findings reflect the differences in task demands between constructing argument and causal maps and provide insights into why and when specific processes produce higher-quality maps. They also offer guidance on developing future diagramming tools and algorithms for automating map analysis and presentation of dynamic support for enhancing student learning, understanding, and problem-solving skills.