The present study addresses the use of one class of representation, >metaphor graphics, as an aid to monitoring a rapidly changing data intensive situation, in this case a patient on a mechanical ventilator. Eight experts in the field of mechanical ventilation viewed summaries containing seven days of data related to a single patient and, as quickly as possible, decided whether the patient was steadily improving, steadily worsening, or was initially moving in one of these two directions but then reversed course. Accuracy was extremely high for both graphic and numerical versions of the summaries but graphic representation led to performance that was significantly faster, averaging half the time required for numeric representation. These findings complement and extend previous research showing that metaphoric graphic representation leads to Bayesian diagnostic reasoning that is extremely accurate but much faster than reasoning via numerical tables.
展开▼
