This paper reports a test of the theory of decision support systems design for user calibration that compares the efficiency of the visual computing paradigm with that of the conventional text paradigm over varied levels of problem novelty. Perfect user calibration exists when a user’s confidence in a decision equals the quality of the decision. The laboratory study reported here compared the effects on user calibration of problems depicted either using a text paradigm or visual computing paradigm. The results support the theory. When problems are new and novel, visual depiction improves user calibration. As problems became more familiar and problem novelty decreases, no difference was found in user calibration between subjects exposed to visibility diagrams and those exposed to a traditional text paradigm.