Angela C. Little

Developmental color perception.

Abstract The purpose of this research is to extend the investigation of surface color perception to several age levels. The 108-item color perception test, originally used with young children, employs six Munsell hue matrices divided into nine combinations each of low, mid, and high chroma and low, mid, and high value across two levels of hue difficulty. The pattern of error results are the same across the age groups comprising kindergartners, fifth graders, high school sophomores, nonartist adults, and professional artists: the lower the chroma and value, the higher the error rate; the higher the chroma and value, the lower the error rate. In hues, green and red are most difficult; orange and yellow are easiest. The frequency of error is linear with respect to age: the younger the group, the higher the error. The latency data differ with respect to age: adults are slowest, followed by kindergartners and fifth graders. High school sophomores are the fastest. The remaining latency results parallel the error results: the lower the chroma and value, the longer the latency and the more difficult the hue, the longer the latency. A set of surface color perception rules are generated.

Evaluation of Single-Number Expressions of Color Difference*

Single-number expressions of color differences among three pairs of samples of vegetable purees were calculated by eleven methods of computing ΔE. The samples differed mainly in chromaticity, and showed color differences more complex than simple nonmetameric ones. Experimental conditions were controlled by limiting the study to an analysis of the relative ability with which the eleven methods were able to rank the three pairs of samples. Serious discrepancies in magnitude and order were found.

Perception of hue re-examined: an analytical consideration of color-oddity test results.

The purpose of this study is to investigate an earlier finding wherein more than 100 subjects in four age groups responded systematically but differently to Munsell hues. According to the theoretical construction of the Munsell Color System, the spacing of the 10 hues is in perceptually equal intervals; the error responses to all hues at constant chroma should therefore be equal. The mean error rates were compared with Munsell hue distribution on seven linear and non-linear transformations of the International Commission on Illumination chromaticity diagram to uniform chromaticity systems. Hue intervals are not equal: red and green have the smallest intervals and largest error, and yellow and yellow-red the largest intervals and smallest error rate. These observations were substantiated by results from multidimensional scaling experiments reported elsewhere.