Fig. 11.

Estimation of illumination of DressO and the Jacket. The first column (a-b) reproduces the bar-type visualizations that we gave observers to indicate their choice along the brightness and yellow-blue gradations. The second and third column show the estimations for the illumination in the background of the dress (c-d) and for the light taken to be shining on the dress (e-f). The last column illustrates the estimations of the light taken to be shining on the Jacket. Blue bars refer to groups with scores below and yellowish bars to groups with scores above zero. BK = Blue-black (negative score for dress), WG = White-gold (positive score for dress), BW = Blue-white (negative score for Jacket), GGo = Green-Gold (positive score for Jacket). Error bars correspond to standard errors of mean. Symbols above the bars report significance of t-tests across observers comparing observers with dress and jacket scores above and below zero. ^***p < 0.001, ns = not significant. For results in the preliminary experiment see Fig. S2. Note that the correlations between perceived dress colours and estimations of the brightness and colour of the illumination of the dress (e-f) replicate the laboratory findings (Witzel et al., 2017). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)