Purchase this article with an account.
Marcelo Costa, Carlo Gaddi; Color Name Distances Scaled by Thurstone's Ranking Order Psychophysical Method. Journal of Vision 2016;16(12):824. doi: 10.1167/16.12.824.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
The psychological continuum of color names or categories is usually studied using categorization methods that have restricted information or statistical multidimensional scaling based on mathematical assumptions hardly controlled. The aim of our study is to scale color name distances using simplest experimental procedure which enables large-scale applications and for different population groups in an interval scale. Eighty two participants (mean age= 22,3 yrs; SD= 1,7) with normal or corrected to normal visual acuity and normal color vision were evaluated. The task consisted in write a column list of colors, one in each line, as fast as possible in a sheet of paper during the experimental period of 20 seconds. Intending to control replicating same-class colors names with one term were accepted. The data were analyzed based on the rank order scaling method proposed by Thurstone, in which the derived proportion of subjective separations was used to calculate subjective distances. The procedure consisted in calculate the frequency with which color A was placed in rank 1 by the N subjects comparing with the frequencies in which the other colors was placed in rank 1. The same comparison was performed by all other rank positions. Summing for the total number of ranks we could measure the probability that B was perceived in a rank order higher than A. Following, we calculated the Z-score to find the distances between probabilities. Our results shown that the color rank position were red>green(0,23)>blue(0,28)>yellow(0,31)>black(0,32)>white(0,50)>purple(0,62)>orange(0,68)>brown(0,72)>pink(0,86). The numbers in the parenthesis are the subjective distances measured. We concluded that rank order is a simple task and produces quantitative interval class data that could be performed for a large group of population allowing studies and comparisons between them using the same method.
Meeting abstract presented at VSS 2016
This PDF is available to Subscribers Only