On the subjective determination of color in biological specimens: the case of the pelage of small mammals

Abstract

In the field of Biological Sciences, the determination of the color of the specimens is of fundamental importance
to answer a wide variety of questions. For a long time, the visual determination of color was the rule.
Subsequently, several reference color charts, such as those of the Munsell Color System, have been developed
and used. As a first approach to the evaluation of the performance of observers in the characterization of the color
of biological specimens under controlled lighting conditions, a psychophysical experiment was carried out
involving the visual determination of color of biological specimens (small mammals of the Lillo Mammal
Collection, Faculty of Natural Sciences and Miguel Lillo Institute, UNT), by 21 observers, using 27 Munsell
color chips preselected as reference. As a first approach, intra- and inter-observer reliability was tested through a
descriptive analysis. The crude results reveal very low intra-observer reliability (with an average percentage of
coincidence in the assignment of the same color chip to the same specimens of 20%). On the other hand,
considering the maximum number of coincidences in the allocation of a color chip per specimen, 36.4% of the
times (on average) different observers assigned the most chosen color chip to the same specimens. These
percentages improve markedly when the condition of detecting differences as small as one step in any of the three
perceptual attributes of color (tone, value, and chroma) is relaxed. In this way, the average percentage of
coincidence in the assignment of equivalent color chips to the same specimens was 68%, while 57.5% of the
times (on average) different observers assigned to the same specimens the equivalent most chosen color chips. Of
the three perceptual attributes, small changes in tone are the most difficult to detect.