Improving decision-making performance through argumentation: An argument-based decision support system to compute with evidence

摘要

While research has shown that argument based systems (ABSs) can be used to improve aspects of individual thinking and learning, relatively few studies have shown that ABSs improve decision performance in real world tasks. In this article, we strive to improve the value-proposition of ABSs for decision makers by showing that individuals can, with minimal training, use a novel ABS called Pendo to improve their ability to predict housing market trends. Pendo helps to weight and aggregate evidence through a computational engine to support evidence-based reasoning, a well-documented deficiency in human decision-making. It also supports individuals in the creation of knowledge artifacts that can be used to solve similar problems in the same domain. An unexpected finding and one of the major contributions of this work is that individual unaided decision-making performance was not predictive of an individual's performance with Pendo, even though the average performance of assisted individuals was higher. We infer that the skills activated when using the tool are substantially different than those enacted to solve the same problem without that tool. We discuss the implications this result has for the design and application of ABSs to decision-making, and possibly other decision support technologies.