Abstract
Computer-generated schematic faces have been proposed as a potentially useful medium for communicating complex data. Two experiments were performed to examine faces in this regard. In the first experiment, subjects performed a clustering task in a nine-dimensional Euclidean space. Subjects performed significantly and substantially better when the points in the nine-space were represented by faces than when they were represented in either of two alternate ways. In the second experiment, subjects performed a paired-associate learning task with faces and with a variety of other stimuli of three different dimensionalities. While faces were found to result in generally superior performance, a number of other interesting effects were also observed.
Get full access to this article
View all access options for this article.
