An Easy Way to Show Memory Color Effects

Abstract

This study proposes and evaluates a simple stimulus display that allows one to measure memory color effects (the effect of object knowledge and memory on color perception). The proposed approach is fast and easy and does not require running an extensive experiment. It shows that memory color effects are robust to minor variations due to a lack of color calibration.

Keywords: cognition; color; experience/learning/expertise; memory; memory colors; objects and features; perception.

Figure 1.

Main stimulus display. The left banana is completely gray (like the background), the right one slightly bluish. Which banana looks gray? If neither looks gray choose the most gray one.

Figure 2.

Control stimuli. Chromatic properties as for stimuli in Figure 1.

Figure 3.

Results of Study 1. Panel a illustrates results for all participants together, Panel b separately for women (pink bars) and men (blue bars). The y axis represents the relative frequency of choosing the bluish over the gray exemplar. Error bars show standard errors of mean, the dotted red line chance probability. Symbols at the bottom of the bars indicate significance in binomial tests for differences from chance probability. Symbols on top of the horizontal bars indicate two-tailed significance in McNemar’s exact tests of differences between disk and banana. *** p < .001; ** p < .01; “ns” nonsignificant.

Figure 4.

Stimuli and results of Study 2. The rows in Panel a show the six stimulus pairs in the order of appearance across the different pages of the survey. The symbols between the two objects of each pair were not part of the stimulus display but have been added to indicate the predictions based on the memory color effect. Panel b shows the relative frequency of choosing answer B over A for each stimulus pair. The order of the stimulus pairs along the x axis corresponds to the order of the stimuli shown in Panel A. Apart from that, format as in Figure 3(a).

Figure 5.

Stimuli and results of supplementary survey of Study 2. Format as in Figure 4.