The use or misuse of three-dimensional graphs to represent lower-dimensional data

Some statisticians hold strong opinions regarding graphs with a 3-D look. However, in experiments little attention has been paid to the effects of adding decorative depth. The performance of subjects on pie charts and bar charts with and without 3-D was evaluated in the present experiment. Subjects were asked to make relative magnitude estimates for different graphs. For pie charts, better performance was observed for 2-D than for 3-D charts. For the bar charts, a more differentiated picture emerged: performance was dependent on the position, height and dimension of the bars. However, 3-D bar charts had the one disadvantage that subjects needed more time to evaluate this type of graph.