Absence Makes the Trust in Causal Models Grow Stronger
Abstract
People prefer complex explanations for complex phenomena, but make better choices when given only the information required. Thus there is a tension between the information people want, and the information they are able to use effectively. However, little is known about how the specific types of information included in causal models influences how people perceive them. We examine how omitting information influences how people reason about causal models, varying whether commonly known or unexpected information is removed (Experiment 1) or which parts of a causal path are omitted (Experiment 2). We find that omitting causal information participants expect to see lowers ratings of trust and other factors, while omitting less commonly known information improves ratings. However, causal paths can be simplified without harming perceptions of diagrams.