Decision making with computer graphics

Computer-driven displays have long been thought to help decision making. But the justification for using these devices in decision-making has been long on intuition and short on quantitative analysis. To see if this intuition was right, we conducted an experiment. Eighteen interested and experienced decision-makers in the inventory control field met for twenty class hours of instruction in advanced inventory control techniques. Near the conclusion of the short course, we measured the participants' decision-making abiJity while using a computer-driven display device. This measurement was compared to their decision-making ability while using information presented to them on paper. To provide a vehicle for the investigation, a simulator was written to apply certain inventory control policies to a hypothetical inventory system handling 34 items. This inventory system faces a randomly derived set of orders, price changes, replenishment of stock and other transactions. The simulator has two sets of input parameters: one governs the distributions of transactions; the other establishes the management policy for inventory control. The simulator investigates twelve distinct policies at one time over a simulation cycle of 24 months. Each participant was given, as examination problems, the statistics resulting from two different simulation runs. Each decision-maker, or decider, made a series of decisions; at the end of each simulation month, he ranked the twelve policies in order of desirability. Each decider used printer output on one problem and graphic display presentation on the other. The statistics from the simulation program were