A human factors study of color notation systems for computer graphics

A human factors experiment was conducted that compares three color notation systems for use in computer graphics. Two of the systems (those in common use) represent colors as triples of real numbers in [0, 1]. The third system is based on natural language color categories in English. It was found that users of the natural language based system were significantly more accurate in specifying colors, despite the coarse granularity of that system as compared to the other two. This demonstrates that giving a user choice from a small set of values that are carefully chosen and based on human factors principles works better than providing a much larger and apparently more flexible set of values that are not based on such principles.